KR20190029524A - System and method for training a robot to autonomously traverse a path - Google Patents

Abstract

A system and method for training a robot to autonomously traverse a path. In one embodiment, the robot can detect an initial placement at an initialization position. Beginning at the initialization position, the robot can generate a map of the navigable path and the surrounding environment during the user control demo of the navigable path. After the demonstration, the robot may later detect the second position at the initialization position and then autonomously navigate the navigable path. Then, the robot can then detect an error associated with the generated map. Methods and systems associated with robots are also disclosed.

Description

System and method for training a robot to autonomously traverse a path

preference

This application claims priority to U.S. Provisional Patent Application No. 15 / 152,425, filed May 11, 2016, which is assigned to the same assignee, and is incorporated herein by reference in its entirety.

Copyright

Part of this disclosure of this patent document contains copyrighted material. The copyright owner does not dispute any copy of the patent or any copy of the patent for the patent disclosure, as indicated in the USPTO's patent file or record, but otherwise retains all rights in the copyright.

[0001] This application is generally concerned, in particular, with a robot system and a method of using the same. In particular, in one aspect, the present disclosure is directed to a system and method for training and operating a robot to autonomously drive a path.

Currently, programming robots can often involve consuming coding that is trying to anticipate or anticipate all the situations a robot can encounter. Not only is this approach costly in terms of time, energy and computer resources, but this approach may also limit the capabilities of the robot. For example, many robots may only be effective in a controlled environment with predictable or predefined conditions. The robot may not be effective in a dynamically changing environment and / or in a new environment where the robot is not specifically programmed. When a robot is programmed with general capabilities, robots can be useful in a variety of tasks, but these tasks may be ineffective or ineffective in certain tasks. On the other hand, robots that are programmed to perform specific tasks effectively and efficiently are limited to these tasks and can not perform other tasks. Similarly, many current robots can require professional technicians and other highly skilled workers to program and manipulate them. This requirement increases robot operation time and cost.

This task is particularly prominent in programming the robot to travel along the path. For example, to program the robot to autonomously travel from a first point to a second point, the programmer may program the map and, depending on the order or logic the robot must travel to that point, Each point on the map should be identified. The programmer may need to program the robot for each environment and enter all the desired paths along with the environment map. Alternatively, if a programmer programs the general rules and logic for a robot to determine a path, then the robot may be slow and inefficient if it follows a particular path. In any case, such programming can be time-consuming and require highly skilled workers to operate the robot. Thus, there is a need for an improved system and method for programming a robot to drive a path.

The foregoing need is met particularly with the present disclosure, which provides a system and method for initializing and operating a robot for driving a trained path. The exemplary implementations described herein have innovative features, which are not necessarily indispensable or cause for their desired attributes. Without limiting the scope of the claims, some of the advantageous features are summarized below.

In some implementations of the present disclosure, the robot can autonomously learn the path by demonstration while navigating and later repeat the demonstrated path.

In a first aspect, a robot is disclosed. In one illustrative embodiment, the robot includes a mapping and localization unit configured to generate a map of the navigable path and environment while demonstrating a navigable path to the robot starting from the initialization position. The robot also includes a navigation unit configured to autonomously navigate the robot using the map.

In one variation, the navigation unit is configured to determine not to autonomously navigate at least a portion of the navigable path.

In another variation, the robot further comprises a sensor unit configured to generate sensor data that at least partially displays objects within the sensor range of the robot, wherein the navigation unit is operable, at least in part, And is further configured to autonomously navigate on a per-view basis.

In another variation, the robot further comprises a first actuator unit configured to actuate the brush. In another variation, the robot further has a second actuator configured to turn the robot.

In another variation, the robot further comprises a processor configured to associate a location on the map with an operation of the unit of the first actuator. In another variation, the robot includes a processor configured to associate a location on the map with an operation of the second actuator unit.

In another variation, the robot includes an interface unit configured to receive a selection of a map generated from a user, the robot autonomously navigates based at least in part on the received selection.

In another variation, the robot has a map evaluation unit configured to correct errors in the map. In another variation, correction of the error includes machine learning that associates at least one of the errors in the map with at least a portion of the corrected map.

In another variation, the robot further comprises a communication unit configured to communicate with a server, which transmits the map to the server and receives a verification of the quality of the map.

In a second aspect, a method of training a robot is disclosed. In one exemplary implementation, the method includes detecting a first position of the robot at an initialization position; Generating a map of the navigable path and the surrounding environment while demonstrating a navigable path to the robot starting from the initialization position; Detecting a second position of the robot at the initialization position; And causing the robot to autonomously navigate at least a portion of the navigable path from the initialization position.

In one variation, the method includes evaluating the generated map for errors; And requesting the user to demonstrate the navigable path back to the robot based at least in part on the errors.

In another variation, the method further comprises correcting errors in the map. In another variation, the method further comprises determining not to autonomously navigate at least a portion of the navigable path.

In another variation, the method further comprises associating the initial location with a map of the navigable path and environment.

In another variation, the method further comprises mapping actions performed by the robot on the navigable path onto the generated map.

In a third aspect, a method of using a robot is disclosed. In one exemplary implementation, the method includes detecting a first position of the robot at an initialization position; Generating a map of the navigable path and the surrounding environment while demonstrating a navigable path to the robot starting from the initialization position; Detecting a second position of the robot at the initialization position; And causing the robot to autonomously navigate at least a portion of the navigable path from the initialization position.

In one variation, the method further comprises associating the initial location with a map of the navigable path and environment.

In another variation, the method further comprises mapping actions performed by the robot on the navigable path onto the generated map.

In a fourth aspect, a non-transitory computer readable medium is disclosed. In an exemplary implementation, a non-transient computer readable storage medium is disclosed wherein a plurality of instructions are stored. The instructions being executable by the processing device to operate the robot, the instructions further comprising instructions for causing the processing device to: detect a first position of the robot at an initialization position; Generating a map of the navigable path and the surrounding environment while demonstrating a navigable path to the robot starting from the initialization position; Detect a second position of the robot at the initialization position; And to cause the robot to autonomously navigate at least a portion of the navigable path from the initialization position.

In one variation, the non-transitory computer-readable storage medium includes instructions that, when executed by the processing device, cause the processing device to: evaluate the generated map for errors; and generate, based at least in part on the errors, Further comprising instructing the robot to request the user to demonstrate the navigable path again.

In another variation, the non-temporal computer-readable storage medium includes instructions that, when executed by the processing device, cause the processing device to correct errors in the map.

In another variation, the non-transitory computer-readable storage medium includes instructions that, when executed by the processing device, cause the processing device to execute instructions that, when autonomously navigating the navigable path, cause the robot to avoid an occasionally placed obstacle To the robot.

In another variation, the non-transitory computer-readable storage medium further comprises instructions, when executed by the processing device, to cause the processing device to determine not to autonomously navigate at least a portion of the navigable path.

In another variation, the non-temporal computer-readable storage medium includes instructions that, when executed by the processing device, cause the processing device to associate the initial location with a map of a navigable path and environment.

In another variation, the generation of a map of the navigable path and the environment further comprises instructions configured to sense the environment to the sensor.

In another variation, the non-transitory computer-readable storage medium further comprises instructions, when executed, to cause the processing device to communicate with a server, the robot transmitting the map to the server, And receives the verification.

In a fifth aspect, an environment and a robot are disclosed. In one illustrative embodiment, the robot includes a mapping and localization unit configured to generate a map of the navigable path and environment while demonstrating a navigable path to the robot starting from the initialization position. The robot also includes a navigation unit configured to autonomously navigate the robot using the map.

In one variation, the navigation unit is further configured to determine not to autonomously navigate at least a portion of the navigable path. This determination includes a determination to avoid obstacles in the environment.

In another variation, the robot further comprises a sensor unit configured to generate sensor data that at least partially displays objects within a sensor range of the robot, wherein the navigation unit is configured to generate sensor data based at least in part on the generated sensor data So as to autonomously navigate the environment.

In another variation, the robot further comprises a first actuator unit configured to actuate a cleaning brush. In another variation, the robot further comprises a second actuator unit configured to turn the robot in an environment.

In another variation, the robot further comprises a processor configured to associate a position on the map with an actuation of the first actuator unit. In another variation, the robot further comprises a processor configured to associate a position on the map with an actuation of the second actuator unit.

Additional aspects and embodiments are disclosed in this disclosure. For example, some implementations include a non-transitory computer-readable storage medium having stored thereon a plurality of instructions, the instructions being executable by a processing device to operate the robot, When executed by the processor, causes the processing device to: detect a first position of the robot at an initialization position; Generating a map of the navigable path and the surrounding environment while demonstrating a navigable path to the robot starting from the initialization position; Detect a second position of the robot at the initialization position; And to cause the robot to autonomously navigate at least a portion of the navigable path from the initialization position.

In some embodiments, when executed by the processing device, the processing device may cause the processing device to evaluate the generated map for errors, and to re-demote the navigable path to the robot based at least in part on the errors To the user. In some implementations, the errors include at least one of a discontinuity of the navigable path in the map and a discontinuity in the environment in the map. In some implementations, the errors include at least duplicate objects. In some implementations, the errors include failures in forming a closed loop. In some implementations, the errors include certain error patterns in the map.

In some implementations, the non-transitory computer readable storage medium further comprises instructions for causing the processing device to correct errors in the map when executed by the processing device. In some implementations, the correction of errors includes machine learning that associates at least one of the errors in the map with at least a portion of the corrected map.

In some embodiments, the non-transitory computer-readable storage medium, when executed by the processing device, causes the processing device to provide instructions to avoid a temporarily placed obstacle while the robot is autonomously navigating the navigable path Command. In some implementations, the non-transitory computer readable storage medium further comprises instructions, when executed by the processing device, to cause the processing device to determine not to autonomously navigate at least a portion of the navigable path. In some implementations, the decision not to autonomously navigate at least a portion of the navigable path includes a determination to avoid an obstacle.

In some embodiments, the robot further comprises instructions configured to cause the processing device to receive a selection of a navigable map from a user interface. In some embodiments, the non-temporal computer-readable storage medium includes instructions that, when executed by the processing device, cause the processing device to associate the initial location with a map of a navigable path and environment.

In some embodiments, having the robot autonomously navigate is configured to cause the processing device to determine the navigable path based at least in part on associating the map of the navigable path and the environment with the initialization position Command. In some embodiments, having the robot autonomously navigate further comprises instructions configured to cause the processing device to navigate based at least in part on the generated map.

In some embodiments, the robot is a floor cleaner. In some embodiments, the robot is a floor scrubber.

In some implementations, the generated map includes an indication of at least a portion of an action performed by the robot on the navigable path. In some implementations, the action is to clean the floor. In some implementations, the behavior is turning.

In some implementations, the generation of the map of the navigable path and the environment further includes instructions configured to sense the environment to the sensor. In some implementations, the generation of the navigable path and the map of the ambient environment further includes instructions configured to sense the ambient environment with the three-dimensional sensor.

In some embodiments, the non-transitory computer readable medium further comprises instructions, when executed, to cause the processing device to communicate with a server, the robot transmitting a map to the server and receiving a verification of the quality of the map do.

As another example, some implementations may include detecting a first position of the robot at an initialization position; Generating a map of the navigable path and the surrounding environment while demonstrating a navigable path to the robot starting from the initialization position; Detecting a second position of the robot at the initialization position; And causing the robot to autonomously navigate at least a portion of the navigable path from the initialization position.

In some implementations, the method includes evaluating the generated map for the errors; And requesting the user to demote the navigable path back to the robot based at least in part on the errors. In some implementations, evaluating the generated map for errors includes identifying overlapping objects. In some implementations, evaluating the generated map for errors includes identifying a failure in forming a closed loop.

In some implementations, evaluating the generated map for errors includes identifying a predetermined pattern in the map. In some implementations, the method further comprises transmitting the map to a server and receiving a signal from the server that at least partially indicates the quality of the map.

In some implementations, the method further comprises correcting errors in the map. In some implementations, correcting errors includes machine learning associating at least one of the errors in the map with at least a portion of the corrected map.

In some implementations, the method further comprises determining to not autonomously navigate at least a portion of the navigable path. In some embodiments, determining not to autonomously navigate at least a portion of the navigable path includes determining to avoid an obstacle.

In some implementations, the demo includes receiving a control signal from a user. In some implementations, the generation of a map of the navigable path and the environment further includes sensing the ambient environment with the sensor. In some embodiments, generating the map of the navigable path and the environment further comprises sensing the ambient environment with a three-dimensional sensor.

In some embodiments, the step of autonomously navigating the robot further comprises receiving a selection of a navigation path from the user interface. In some embodiments, the step of allowing the robot to autonomously navigate comprises navigating using a map of the navigable path and the environment.

In some implementations, the method further comprises associating the map of the navigable path and the environment with the initialization location.

In some implementations, the method further comprises determining the navigable path based at least in part on associating the map of the navigable path and the environment to the initialization position.

In some implementations, the method further comprises mapping actions performed by the robot on a navigable path onto the generated map. In some implementations, the action includes cleaning the floor. In some implementations, the behavior includes turning.

As another example, some implementations may include a non-transitory computer readable storage medium having stored thereon a plurality of instructions, the instructions being executable by a processing device to operate the robot, , The processing device is configured to: generate a map of the navigable path and the surrounding environment while demonstrating the navigable path to the robot starting from the initialization position.

In some implementations, the generated map further includes an indication indicating at least a portion of an action performed by the robot on the navigable path. In some implementations, the action is to clean the floor. In some embodiments, the robot is a floor cleaner. In some implementations, the demo of the navigation path is a computer simulation.

As another example, some implementations may include a mapping and localization unit configured to generate a map of the navigable path and environment while demonstrating a navigable path to the robot starting from the initialization position; And a robot including a navigation unit configured to autonomously navigate the robot using the map.

In some implementations, the navigation unit is further configured to determine not to autonomously navigate at least a portion of the navigable path. In some implementations, the decision not to autonomously navigate includes a determination to avoid obstacles in the environment.

In some embodiments, the robot further comprises a sensor unit configured to generate sensor data that at least partially displays objects within a sensor range of the robot, wherein the navigation unit is configured to generate sensor data based at least in part on the generated sensor data, So as to autonomously navigate.

In some embodiments, the robot further comprises a first actuator unit configured to actuate the brush. In some embodiments, the robot further includes a second actuator unit configured to turn the robot.

In some embodiments, the robot further comprises a processor configured to associate a position on the map with an actuation of the first actuator unit. In some embodiments, the robot further comprises a processor configured to associate a location on the map with an operation of the second actuator unit.

In some implementations, the robot further comprises a user interface configured to receive a selection of the generated map from the user, and the robot autonomously navigates based at least in part on the received selection.

In some implementations, the robot further includes a map evaluation unit configured to correct errors in the map.

In some implementations, the correction of the error includes machine learning that associates at least one of the errors in the map with at least a portion of the corrected map.

In some implementations, the errors include at least a superposition object. In some implementations, the errors include failures in forming a closed loop. In some implementations, the errors include certain patterns in the map.

In some implementations, the map evaluation unit is further configured to correct errors in the map. In some implementations, the correction of errors includes machine learning that associates at least one of the errors in the map with at least a portion of the corrected map.

In some implementations, the processor is further configured to associate a map of the navigable path and the environment with the initialization location.

In some implementations, the processor is further configured to determine the navigable path based at least in part on an association of the navigable path and a map of the environment to the initialization position. In some embodiments, the navigation unit further configured to autonomously navigate the robot further comprises instructions configured to cause the processing device to navigate based at least in part on the generated map.

In some embodiments, the robot further comprises a communication unit configured to communicate with the server, the robot transmitting the map to the server and receiving verification of the displayed quality of the map.

BRIEF DESCRIPTION OF THE DRAWINGS These and other objects, features and characteristics of the present disclosure, as well as combinations of components and economies of operation and functionality of the operation and function of the related elements of the structure, will become more apparent when taken in conjunction with the following description and the appended claims It will become clear. Reference is made to the appended claims, all of which form a part hereof, wherein like reference numerals designate corresponding parts throughout the several views. It is to be expressly understood, however, that the drawings are for purposes of illustration and description only and are not intended as a definition of the limits of the disclosure. As used in the specification and claims, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise .

BRIEF DESCRIPTION OF THE DRAWINGS In the following, aspects disclosed with reference to the accompanying drawings, provided so as not to limit the disclosed aspects for purposes of explanation, are designated by the same names.
FIG. 1A is an overhead view of one exemplary path that is autonomously navigated by a robot, in accordance with an implementation of the present disclosure. FIG.
1B is an overhead view of the exemplary path shown in FIG. 1A, illustrating a user demonstrating a path to a robot, in accordance with an implementation of the present disclosure.
FIG. 1C is an overhead view of an alternative exemplary path that is autonomously navigated by the robot shown in FIGS. 1A and 1B, and the robot avoids the object in accordance with the principles of the present disclosure.
2 is a process flow diagram of an exemplary method for training a robot to autonomously navigate an exemplary path, in accordance with the principles of the present disclosure;
Figure 3 is a functional block diagram of an exemplary robot, in accordance with some embodiments of the present disclosure.
4 is a process flow diagram of an exemplary method of running after an exemplary robot has learned an exemplary path, in accordance with the principles of the present disclosure;
5A is an exemplary user interface for receiving input from a user to teach or select an exemplary path, in accordance with the principles of the present disclosure.
Figures 5B-D are overhead views of an exemplary robot for detecting an initialization position and initializing an exemplary orientation and exemplary position, in accordance with the principles of the present disclosure;
5E is an overhead view of an exemplary robot in which the robot emits an energy pattern, in accordance with the principles of the present disclosure;
6A is a side view illustrating a user controlling a robot while demonstrating an autonomous navigation path to the robot, in accordance with the principles of the present disclosure;
6B illustrates various side views of an exemplary body shape for a floor scrubber, in accordance with the principles of the present disclosure;
Figure 6C illustrates various side views of an exemplary body shape of a robot, in accordance with the principles of the present disclosure;
6D is an overhead view of a user controlling the robot while sensing the surroundings, in accordance with the principles of the present disclosure;
Figures 7A and 7B illustrate various exemplary maps generated by a robot as the robot travels in an environment, in accordance with the principles of the present disclosure.
Figures 8a and 8b illustrate various exemplary mapped objects that may appear in a map, Figure 8a illustrates one set of exemplary objects substantially parallel to each other, while Figure 8b illustrates a set of exemplary objects, And another set of exemplary objects that are not substantially parallel to one another.
Figure 8C is an overhead view of a mask used to retrieve a map for a substantially parallel object, in accordance with the principles of the present disclosure.
9A is an overhead view of an exemplary path discontinuity between path portions of a map, in accordance with the principles of the present disclosure;
9B is an overhead view of an object discontinuity between object portions of the map, in accordance with the principles of the present disclosure;
9C is an overhead view of a mapping portion having an exemplary discontinuity including both a path discontinuity and an object discontinuity, in accordance with the principles of the present disclosure;
10 is an overhead view of a mapped portion having an exemplary overlapping object, in accordance with the principles of the present disclosure;
11A is an overhead view of a robot running in an exemplary closed-loop path, in accordance with the principles of the present disclosure, wherein the exemplary initialization position is substantially similar to an exemplary ending position.
11B is an exemplary mapping error that relates the mapping error to the corrected path of the robot, in accordance with the principles of the present disclosure
Figure 12 is an exemplary user interface that may be used for path selection, in accordance with the principles of this disclosure.
13 is a process flow diagram of an exemplary method of operating a robot, in accordance with the principles of the present disclosure;
All drawings herein are © Copyright 2017 Brain Corporation All rights reserved.

Next, various aspects of the novel systems, apparatus, and methods disclosed herein will be described in more detail with reference to the accompanying drawings. This disclosure, however, may be embodied in many different forms and should not be construed as limited to any specific structure or function presented throughout this disclosure. Rather, these aspects are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. Based on the teachings herein, those skilled in the art will readily appreciate that any scope of the present disclosure, whether embodied independently or in combination with any other aspects of the disclosure, may be embodied in any of the novel systems, And are intended to be inclusive. For example, an apparatus may be implemented using any number of aspects set forth herein and methods may be practiced. Also, the scope of this disclosure is intended to cover such apparatus or methods that are implemented using other structure, functionality, or structure and functionality in addition to or in addition to the various aspects of the disclosure described herein. It is to be understood that any aspect of the disclosure described herein may be embodied by one or more elements of the claims.

While certain implementations have been described herein, many variations and permutations of these implementations fall within the scope of the present disclosure. While certain advantages and advantages of these embodiments are mentioned, the scope of the present disclosure is not intended to be limited to any particular advantage, use, and / or purpose. The description and drawings are merely illustrative of the present disclosure rather than limiting, and the scope of the present disclosure is defined by the appended claims and their equivalents.

The present disclosure provides an improved system and method for operating a robot for autonomous navigation. As used herein, a robot may comprise a mechanical or virtual entity configured to automatically perform a complex series of actions. In some cases, the robot may be an electro-mechanical machine that is guided by a computer program or electronic circuitry. In some cases, the robot may include an electromechanical machine configured for autonomous navigation, and the robot may move from one position to another, with little or no user control. Such autonomous navigation robots may include autonomous vehicles, floor cleaners (e.g., floor scrubbers, vacuum cleaners, etc.), rovers, drones, and the like. In some implementations, some of the systems and methods described in this disclosure may be implemented in a virtual environment, and a virtual robot may learn a demo path in a simulated environment (e.g., in a computer simulation) can do. After learning these paths, the systems and methods described in this disclosure can be used to autonomously navigate the learned paths in a simulated environment and / or in the real world.

The following provides a detailed description of various embodiments and variations of the systems and methods of this disclosure. Although primarily described in the context of a robot floor cleaner, the systems and methods described herein can be used with other robots, including, for example, any autonomously navigating robots. Those skilled in the art having the content of this disclosure will readily be able to anticipate numerous other exemplary embodiments or uses of the techniques described herein.

Preferably, the system and method of the present disclosure comprise at least: (i) reducing the need for environment specific programming; (ii) reduce the need for a highly skilled technician to program the robot; (iii) provide custom performance from a generally programmed robot; (iv) eliminate or reduce the need for task-specific programming (e.g., how close to approaching obstacles for cleaning); And (v) enable effective autonomous navigation of the robot. Other advantages are readily apparent to those skilled in the art having the benefit of the present disclosure.

For example, by training a robot to traverse a path by way of a demo, the user does not have to program all the paths in advance. Preferably, this enables the user to train the robot to navigate an environment that the user did not expect in advance. In addition, the user does not need to utilize any specific expertise to train the robot. For example, a user does not need to be trained in knowing computer science and / or programming robots. Instead, you just know how the robot performs the tasks you want to do. For example, if the robot is a floor cleaner, you just need to know how to clean the floor and demonstrate to the robot.

In some cases, training the robot to travel along the path allows the robot to perform specific tasks on a specification without identifying and programming in each specification. As an example, if the robot is a floor scrubbing unit, it may be desirable for the floor cleaner unit to drive a distance from a wall, shelf, and the like. Users can demonstrate these distances when training robots, and in some cases robots can repeat these distances.

Moreover, training robots that can learn navigable paths allows robots to be programmed generally to perform in a number of environments while specifically programming to efficiently navigate a particular environment. Advantageously, this allows the robot to be optimized for a particular application and at the same time have the advantage of having the ability and flexibility to perform in various applications.

In some implementations, the maps and paths may be verified and / or approved prior to navigation. This verification and / or validation can prevent accidents and / or situations in which the robot can collide with the walls and / or obstacles due to poor map and / or path quality.

FIG. 1A illustrates an overhead view of an exemplary path 106 autonomously navigated by a robot 102 through an implementation of this disclosure. The robot 102 may autonomously navigate through the environment 100, which may include various objects 108, 110, 112, 118. The robot 102 may start at the initialization position 104 and end at the ending position 114. [

By way of illustration, in some implementations, the robot 102 may be a robot floor cleaner, such as a robot floor scrubber, vacuum cleaner, steamer, mop, sweeper, and the like. The environment 100 may be a space having a floor to be cleaned. For example, the environment 100 may be a store, a warehouse, an office building, a house, a storage facility, and the like. One or more of the objects 108, 110, 112, 118 may be a shelf, display, object, item, person, animal or any other entity that may be on the floor or interfere with the robot's ability to navigate through the environment Or something. The path 106 may be a cleaning path traveled by the robot 102. The path 106 may follow a weave path between the objects 108, 110, 112, 118 as shown in the exemplary path 106. For example, if the objects 108, 110, 112, 118 are shelves of a store, the robot 102 may clean the floor of the hallway along the corridor of the store. However, other routes may also be contemplated, such as unobstructed floor areas and / or weaving back and forth along any cleaning path that a user may use to clean the floor, without limitation. Thus, one or more of the paths 106, 116, 126 illustrated in FIGS. 1A, 1B, and 1C, respectively, may appear differently, as illustrated, and are only meant by way of example. As illustrated, an example of an environment 100 is shown, but an environment 100 can take any number of forms and arrangements (e.g., of any size, configuration, and layout of a room or building) And is not limited by this disclosure.

In path 106, the robot 102 may clean along the path 106, starting at an initialization position 104, which may be its starting point, until it reaches an ending position 114 where cleaning can be stopped . End position 114 may be designated by user 604 as described with reference to Fig. 6A. In some cases, the end position 114 may be a position in the path 106 where the robot 102 then sweeps the desired area of the floor. In some cases, the end position 114 may be the same or substantially similar to the initial position 104 so that the robot 102 performs a substantially closed loop of cleaning and terminates at its starting point, i.e., . In some cases, the end position 114 may be a position for storing the robot 102, such as a temporary parking place, a storage room, or a closet. In some cases, the end position 114 may be the point at which the user 604 decides to clean and stop training the robot 102. The robot 102 may or may not clean at all points along the path 106. For example, when the robot 102 is a robot floor scrubber, the cleaning system (e.g., water flow, cleaning brush, etc.) of the robot 102 may be moved Thus moving in a particular order), but may not operate in other parts and / or in some trajectories. This may be desirable when only some areas of the floor are cleaned and other areas are not cleaned. In this case, the robot 102 may turn on the cleaning system in the area that the user 604 has demonstrated to the robot 102, or else it may turn off the cleaning system.

1B illustrates an overhead view of a user 604 demonstrating path 116 to robot 102 before robot 102 autonomously traverses path 106 in environment 100. FIG. When demoting the path 116, the user may start the robot 102 at the initialization position 104. The robot 102 may then weave around the objects 108, 110, 112, 118. The robot 102 may ultimately be terminated at the end position 114. In some cases, the autonomously navigated path 106 may be exactly the same as the demoted path 116. In some cases, the path 106 may not be exactly the same as the path 116, but may be substantially similar. For example, as the robot 102 navigates the path 106, the robot 102 may be moved to sensors (e.g., as described with reference to Figures 5B-5E) to sense whether the robot 102 is related to its surroundings Sensors such as sensors 560A-D and / or sensors 568A-B). This sensing may not be accurate in some cases, causing the robot 102 to demonstrate and not to navigate the correct path 102 trained to follow by the robot 102. In some cases, a small change in the environment 100, such as a movement of the shelf and / or a change in items on the shelf, may cause the robot 102 to move along the path (s) when the robot autonomously navigates the path 106 116). As another example, as illustrated in Fig. 1C, the robot 102 may autonomously route a path 126, which may be another path traveled by the robot 102, based at least in part on the demoted path 116 Objects 130 and 132 can be bypassed and avoided during navigation. Objects 130 and 132 would not exist (and not be avoided) when the user demohed path 116. [ For example, objects 130 and 132 may be changes and / or dynamic changes to an object / item that is temporarily placed and / or changed, and / or the environment. As another example, the user 604 may have made a mistake when demoting the path 116. For example, the user 604 may have collided or collided with a wall, shelf, object, obstacle, or the like. In such a case, the robot 102 may store one or more behaviors that the robot can correct, such as collisions and / or collisions with walls, shelves, objects, obstacles, etc., to memory (e.g., memory 302) have. Thereafter, while the robot 102 is autonomously navigating the demoted path 116 as the path 126, the robot 102 may modify these behaviors when autonomously navigating and / or do not perform them (e.g., For example, collide and / or collide with walls, shelves, objects, obstacles, etc.). As such, the robot 102 may determine not to autonomously navigate at least a portion of the navigable path, such as the demo path. In some implementations, deciding not to autonomously navigate at least a portion of the navigable path includes determining when to avoid obstacles and / or objects.

As previously mentioned, as the user demonstrates path 116, the user can determine where to clean (e.g., at what location) and / or along a path along path 116 (E.g., when the robot 102 autonomously cleans the path 106), the cleaning system of the robot 102 may be turned on and off or other actions may be performed. The robot can perform these actions when writing these actions in the memory 302 and then autonomously navigating. These actions include turning, turning the water on and off, splashing water, turning the vacuum on or off, moving the vacuum hose position, moving the arms, lifting and lowering the lift, moving the sensor, turning the sensor on and off , And any actions that the robot 102 can perform.

FIG. 2 illustrates a process flow diagram of an exemplary method 200 for training a robot 102 to autonomously navigate a path 106. The portion 202 includes locating the robot 102 in the initialization position 104. This first arrangement of the robots 102 to the initialization position 104 may be used to control the robot 102 to move to the initialization position 104 or to remotely manipulate, push, or otherwise control the robot 102, May be performed by a user 604 (described later with reference to Fig. 6), which may be a person. For example, the user 604 may send a control signal to the robot 102. The robot 102 may receive these control signals as commands for movement.

Returning to Fig. 2, the portion 204 includes demonstrating the navigation path 116 to the robot 102. 1B, the user 604 may demonstrate to the robot 102, without limitation, by driving, remotely manipulating, pushing, or otherwise controlling the robot 102 along path 116 have. For example, the user 604 may transmit a control signal to the robot 102. [ The robot 102 may receive these control signals as commands for movement. A plurality of such movements together can form a demo path. In this way, the user 604 can demonstrate the desired travel path to the robot 102. [ In the context of a robotic floor cleaner, the demoted path 116 may be a desired path to clean the floor. In this manner, the user 604 can train the robot to learn how to clean the floor.

Returning to Fig. 2, the portion 206 includes placing the robot 102 in the initialization position 104 once again. This second configuration of the robot 102 to the initialization position 104 may be performed at a later point in time after the portion 204, such as immediately after the demonstration of the portion 204, or a few hours, a few days, Such as whenever the user 604 wishes to clean the floor.

Returning to Fig. 2, portion 208 includes initiating autonomous navigation. In some cases, after the user initiates autonomous navigation, the robot 102 may travel along a path 106 (or path 126 in some cases) that may be substantially similar to the demoted path 116 have. In some implementations, the user 604 may select a demo path on the user interface 318 as described with reference to FIG. 11A. As an example using FIG. 1A, the robot 102 may then autonomously navigate a path substantially similar to the path or path 106 from the initialization position to the end position 114.

FIG. 3 illustrates, in some implementations, a functional block diagram of an exemplary robot 102. 3, the robot 102 includes a controller 304, a memory 302, a power supply 306, and an operation unit 308, each of which may be coupled to each other, 0.0 > and / or < / RTI > communicatively coupled to the subcomponent. The controller 304 controls various operations performed by the robot 102. Although a particular implementation is illustrated in FIG. 3, it will be appreciated that the architecture may be modified in some implementations, as will be readily apparent to those skilled in the art having the benefit of the present disclosure.

Controller 304 may include one or more processors (e.g., microprocessors) and other peripheral devices. As used herein, a processor, microprocessor, and digital processor may include, without limitation, a digital signal processor ("DSPs"), a reduced instruction set computer ("RISC"), a general purpose ("CISC" ("FPGAs"), programmable logic devices ("PLDs"), reconfigurable computer fabrics ("RCFs"), array processors, secure microprocessors, and application specific integrated circuits And the like. Such a digital processor may be included on a single unitary integrated circuit die or may be distributed across multiple components.

The controller 304 may be operably and / or communicatively coupled to the memory 302. The memory 302 may be, without limitation, a read only memory ("ROM"), a random access memory ("RAM"), a nonvolatile random access memory ("NVRAM"), a programmable read only memory Erasable programmable read only memory ("EEPROM"), dynamic random-access memory (DRAM), mobile DRAM, synchronous DRAM ("SDRAM"), dual data rate SDRAM (DDR / SDRAM) Quot;), fast page mode RAM ("FPM"), latency decreasing DRAM ("RLDRAM"), static RAM ("SRAM"), flash memory (e.g., NAND / NOR) RAM ("PSRAM"), and the like. Memory 302 may provide commands and data to controller 304. For example, memory 302 may be a non-transitory computer readable storage medium that stores a plurality of instructions executable by a processing device (e.g., controller 304) to operate robot 102. In some cases, the instructions may be configured to cause the processing device to perform the various methods, features, and / or functions described in this disclosure when executed by the processing device. Thus, the controller 304 may perform logical and arithmetic operations based on program instructions stored in the memory 302. [

The operation unit 308 may be coupled to the controller 304 or other controller to perform the various operations described in this disclosure. In some implementations, one or more of the modules in operational units 308 may be included, or none of them may be included. Throughout this disclosure, various controllers and / or processors are referenced. In some implementations, a single controller (e.g., controller 304) may function as the various controllers and / or processors described. In other implementations, different controllers and / or processors may be used, such as controllers and / or processors specifically used in one or more of the operational units 308. [ The controller 304 sends a signal such as a power signal, a control signal, a sensor signal, a query signal, a status signal, a data signal, an electrical signal, and / or any other desired signal, including discrete and analog signals, ≪ / RTI > Controller 304 may adjust and / or manage operation unit 308 and / or may set timing (e.g., synchronously or asynchronously), turn on or off, Or transmit / receive data to / from the robot 102, and / or receive, transmit, and / or update network parameters, update firmware, update query information, Can perform some operations to execute.

The operation unit 308 may include various units that perform functions for the robot 102. For example, the units of operation unit 308 may include a mapping and localization unit 312, a sensor unit 314, a map evaluation unit 324, an actuator unit 318, a communication unit 316, a navigation unit 326, And a user interface unit 322. The operation unit 308 may also include other units that provide various functionalities of the robot. In some cases, the units of operation unit 308 may be instantiated in software or in hardware or in software and hardware. For example, in some cases, units of operation unit 308 may include computer-implemented instructions that are executed by a controller. In some cases, the units of operation unit 308 may include hard-coded logic. For example, in some cases, the units of the operational unit 308 may include both computer-executable instructions executed by the controller, and hard-coded logic. When the operating unit 308 is at least partially implemented in software, the operating unit 308 may include units / modules of code that are configured to provide one or more functionality.

In some implementations, the sensor unit 314 may include a system that is capable of detecting features within and / or around the robot 102. The sensor unit 314 may include sensors that are internal or external to the robot 102 and / or may have components that are partially internal and / or external to the part. The sensor unit 314 may comprise an external water soluble sensor such as a sonar, a radar, a radar, a laser, a video camera, an infrared camera, a 3D sensor, a 3D camera, and / or any other sensor known in the art. The sensor unit 314 may also include a proprioceptive sensor such as an accelerometer, an inertial measurement unit, a mileage meter, a gyroscope, a speedometer, and the like. In some implementations, the sensor unit 314 may be configured to measure raw measurements (e.g., current, voltage, resistance, gate logic, etc.) and / or transformed measurements (e.g., distances, angles, Can be collected.

In some implementations, the mapping and localization unit 312 may map the environment 700 (as will be described with reference to Figures 7a-7b) as the robot 102 navigates the environment 100 (or any other environment) (Or any other generated map of any environment) that can be configured and updated in a computer-readable manner. At the same time, the mapping and localization unit 312 may record the demoted path (e.g., path 116) in the map 700. The mapping and localization unit 312 may map the environment 100 and localize (e.g., locate) the robot 102 in the map 700. ("2D"), three-dimensional ("3D"), and / or four-dimensional ("4D") data representing at least in part the environment 100, at least partially by data obtained at least in part by the sensor unit 314, ) Map. ≪ / RTI > For example, the map 700 may include a description that at least partially represents an obstacle and / or an object detected by the robot 102. The map 700 may also record the demoted path, such as the mapped path 716, as described with reference to Figs. 7A and 7B. For example, the mapped path 716 may be located at least partially in the relative position of the robot 102 (e.g., including one or more of position, displacement, and orientation) relative to the same reference as the initialization position 104 (E.g., x and y in a 2D map and x, y and z in a 3D map). This coordinate may include the direction (e.g., displacement angle) of the robot 102 at any given point relative to the reference, such as the initialization position 104. As used herein, the term position has a conventional and customary meaning. For example, in some cases, the location may include locations in terms of objects, displacements, coordinates, etc. of the robot 102, and the like. In some cases, the location may include the direction of the object, the robot 102, and the like. Thus, in some cases, the position and posture may be used interchangeably to include position, displacement, and orientation. The map 700 generated through the demo process may substantially record the entire environment in which the robot 102 sensed in one or more demos / drills. For this reason, a part of the map 700 may be referred to as a global map. In some cases, after the demo, the map 700 may be static in that the map 700 is not substantially updated. In some implementations, the path 716 mapped to the map 700 may also be separately generated (e.g., by a user using a computer) and uploaded onto the robot 102.

The mapping and localization unit 312 may also receive sensor data from the sensor unit 314 to localize (e.g., position) the robot 102 to the map 700. In some implementations, the mapping and localization unit 312 may include a localization system and method that allows the robot 102 to localize itself in the coordinates of the map 700. [ Based at least in part on the data from sensor 314, the mapping and localization unit 312 can infer the location of the robot 102 in the coordinates of the map 700 of the environment 100. The ability to localize the robot 102 with the coordinates of the map 700 allows the robot 102 to navigate the environment using the map 700 to determine the location on the path 716 on which the robot 102 is mapped Can be approximated.

In some implementations, communication unit 316 may include one or more receivers, transmitters, and / or transceivers. The communication unit 316 may be any of a variety of communication devices such as BLUETOOTH®, ZIGBEE®, Wi-Fi, inductive wireless data transmission, radio frequency, wireless transmission, RFID (Radio Frequency Identification), NFC (HSDPA), High Speed Uplink Packet Access (HSUPA), Time Division Multiple Access ("TDMA "), ("FSS"), direct sequence spread spectrum ("DSSS"), and code division multiple access ("CDMA" (PAN / 802.15), WiMAX (worldwide interoperability for microwave access), 802.20, Long Term Evolution (LTE) (e.g., LTE / LTE- A), time division LTE (Time Division Multiple Access), global system for mobile communication (GSM), narrow band / frequency division multiple access ("FDMA"), ("IrDA "), and / or the like, and / or any combination thereof, for example, Or any other form of wireless data transmission.

As used herein, network interfaces include, but are not limited to, FireWire (e.g., FW400, FW800, FWS800, FWS1600, FWS3200, etc.), USB (e.g., USB 1.X, USB 2.0, USB 3.0, USB Type- (E.g., 10/100, 10/100/1000 (gigabit Ethernet), multimedia over coaxial technology (MoCA), Coaxsys (e.g. TVNETTM), radio frequency tuners (e.g., inband or 00B, cable modem, etc.) (E.g., Wi-Fi (802.11), WiMAX (e.g., WiMAX (802.16)), PAN (e.g., PAN / 802.15), cellular (3G, LTE / LTE-A / TD- Data, or software interface with a component, a network, or a processor, including, but not limited to, an IEEE-Std. 802.11, an IEEE- Std. 802.11, IEEE-Std. 802.11 (eg 802.11 a / b / g / n / ac / ad / af / ah / ai / aj / aq / ax / ay) and / can do.

The communication unit 316 may also be configured to transmit / receive a transmission protocol over a wired connection, such as any cable having a signal line and ground. For example, such a cable may include an Ethernet cable, a coaxial cable, a Universal Serial Bus (USB), FireWire and / or any connection known in the art. Such protocols may be used by communication unit 316 to communicate with external systems such as computers, smart phones, tablets, data capture systems, mobile communication networks, clouds, servers, and the like. The communication unit 316 may be configured to transmit and receive signals including numbers, letters, alphanumeric characters and / or symbols. In some cases, encryption is performed using other encryption algorithms that conform to standards such as 128-bit or 256-bit keys, and / or standards such as Advanced Encryption Standard (AES), RSA, Data Encryption Standard can do. Communications 316 may be configured to transmit and receive status, commands and other data / information. For example, the communication unit 316 may communicate with the user controller to allow the user to control the robot 102. [ The communication unit 316 may communicate with the server / network such that the robot 102 may transmit data, status, commands, and other communications to the server. The server may be communicatively coupled to the computer (s) and / or device (s) that may be used to monitor and / or control the robot 102 remotely. The communication unit 316 may also receive updates (e.g., firmware or data updates), data, status, commands and other communications from the server to the robot 102 and / or its operational unit 308.

In some embodiments, the actuator unit 318 may be an electric motor, a gas motor, a driven magnet system, a solenoid / ratchet system, a piezoelectric system (e.g., an inch worm motor), magnetic limiting elements, Or any manner of driving an actuator known in the art. By way of example, and not limitation, such actuators may be wheel or other displaceable enabling drivers (e.g., mechanical legs, jet engines, propellers, hydraulics (not shown), etc.) to enable the robot 102 to navigate through the environment 100, Device, etc.). In some cases, the actuator unit 318 may be used to mobilizing the brush for actions and / or floor cleaning, moving the squeegee (e.g., up, down, left, right, forward, Turning on / off the water, turning the vacuum on / off, moving the vacuum hose position, moving the arms, lifting and lowering the lift, turning on the camera and / or any sensor of the sensor unit 314, and / Or any movement for which the robot 102 wishes to perform an action, such as for movement-specific tasks.

In some implementations, the user interface unit 322 may be configured such that a user (e.g., user 604 or any other user) may interact with the robot 102. For example, the user interface unit 322 may be connected wirelessly or wired (including, without limitation, any of the wireless or wired connections described in this disclosure, as referenced to the communication unit 316) (For example, USB, DVI, Display Port, E-Sata, Firewire, PS / 2, Serial, VGA, SCSI, Audio Port, HDMI, PCMCIA port, A memory card port (e.g., SD and miniSD), and / or a port for computer readable media), a mouse, a roller ball, a console, a vibrator, an audio converter, and / / RTI > and / or < / RTI > The user interface unit 322 may be, without limitation, an LCD, an LED display, an LED LCD display, an IPS display, a cathode ray tube, a plasma display, an HD panel, a 4K display, a retina display, an organic LED display, , Canvas, screen ink technology, and / or a display such as any display, television, monitor, panel, and / or device known in the art of visual presentation. In some implementations, the user interface unit 322 may be located on the body of the robot 102. In some implementations, user interface unit 322 may be located remotely from the body of robot 102, but may be located directly or indirectly (e.g., via a network or cloud) (e.g., via communication unit 316) To the robot 102. The robot 102,

In some implementations, the map evaluation unit 324 may include a comparator, a signal processor, an image processor, and other software or hardware components. As described with reference to Figs. 7A and 7B, Figs. 8A to 8C, 9A to 9C, 10 and 11, the map evaluation unit 324 detects mapping errors, In order to determine the usefulness of the map 700 for autonomous navigation and / or autonomous navigation (e.g., high, good, acceptable, low, and / ) Can be analyzed and evaluated. In some cases, upon analyzing the map 700 or any other map, the map evaluation unit 324 may determine that there were no mapping errors and / or that the map is of poor quality. As a result, the robot 102 may again use a user interface unit 322 or a communication unit 316 to send a path (e.g., path 116) back to a user (e.g., user 604) Demonstrate or prompt to remap environment 100.

In some implementations, the navigation unit 326 may comprise components and / or software configured to provide directional instructions for the robot 102 to navigate. The navigation unit 326 may process map and localization information generated by the mapping and localization unit 312, sensor data from the sensor unit 314 and / or other operational units 308. For example, the navigation unit 326 may receive the map 700 from the mapping and localization unit 312. The navigation unit 326 may also receive localization information from the mapping and localization unit 312 that may at least partially indicate the location of the robot 102 within the map 700, including the path 716. The navigation unit 326 may also receive sensor data from a sensor unit 314 that may at least partially display an object around the robot 102. The navigation unit 326 may instruct the robot 102 to navigate to (e.g., front, left, right, rear, etc.) using one or more of a map, position, and sensor data.

In some embodiments, the power supply 306 may include, but is not limited to, lithium, lithium, nickel-cadmium, nickel-metal hydride, nickel-hydrogen, carbon-zinc, silver oxide, zinc- , Alkaline, or any other type of battery known in the art. Certain batteries may be rechargeable, such as by plugging them wirelessly and / or to an external power source (e.g., by a resonant circuit and / or EEH by a resonant tank circuit). The power supply 306 may also include a wall socket and an electronic device for converting sun, wind, water, nuclear, hydrogen, gasoline, natural gas, fossil fuel, mechanical energy, steam, It can be any energy supply.

In some implementations, the operating system 310 may include a memory 302, a controller 304, a power supply 306, modules in the operational unit 308, and / or any software, hardware, and / And may be configured to manage features. For example and without limitation, the operating system 310 may include a device driver that manages hardware resources for the robot 102.

As noted above, any of the above-described components of the robot 102 may be instantiated with software and / or hardware. For example, the unit / module may be hardware and / or code on a computer.

4 illustrates a process flow diagram of an exemplary method 400 in which the robot 102 learns and traverses a path. For example, at portions 402, 404, 406 in the teaching step 414, the robot 102 may learn the path 116 demoed by the user 604. Subsequently, at portions 408, 416, 412 in the autonomous step 416, the robot 102 may autonomously navigate along the path 106, 126.

In some implementations, the robot 102 may begin the teaching step 4214 by receiving input from the input 574 in the user interface 500 illustrated in FIG. 5A. The user interface 500 may appear on the display 576, which may be a mobile device, a special device, any other device that has a screen and is configured to receive user input. In some cases, the display 576 may be part of the user interface unit 322 of the robot 102. In some cases, the display 576 may be a separate display communicatively coupled to the robot 102, such as, but not limited to, communicatively coupled through the communication unit 316 of the robot 102 . Input 574 may include buttons, wireless buttons, pull-down menus, text input, and / or any manner in which a user may enter information and / or commands known in the art. The user interface 500 may also include an input 572 that may be used to initiate an autonomous step 416, which will be described later in this disclosure. Input 572 may include buttons, wireless buttons, pull-down menus, text input, or any manner in which the user enters information and / or commands known in the art.

Returning to Figure 4, at portion 402, the robot 102 may detect the initialization position 104 and initialize the position and / or orientation of the robot 102. In some implementations, the initialization location 104 is a location for a floor and / or floor plan. For example, the initialization position 104 may be used by the user to indicate that the boundary is to be displayed (e. G., When the robot 102 calls for a learned path) Physically or digitally < / RTI > In some implementations, based on, at least in part, where the user has stopped the robot 102, the robot 102 may detect that the robot 102 is in the initialization position 104 itself. In this way, it is possible to estimate where the robot 102 begins to train in the initialization position 104 after the user has stopped (as described with reference to portion 404). In some implementations, a transmitter (e.g., RFID, NFC, BLUETOOTH®, wireless transmission, radio frequency domain, and / or transmitter that transmits communications using other communication protocols described in this disclosure) , Or may be located substantially adjacent thereto. The robot 102 can detect that the robot 102 itself is in the initialization position 104 when the robot 102 detects that it is in the upper or substantially close vicinity of the transmitter. In some cases, the transmitter may have an operable range that allows it to detect communications from the transmitter only when the robot 102 is in the starting position. As an illustrative example, the transmission range of the NFC may be less than 10 centimeters. Therefore, when the robot 102 receives the transmission via the NFC, the robot 102 can detect that it is located at the initialization position 104. [ In some implementations, the robot 102 can receive transmissions from a transmitter and calculate the distance to the transmitter based at least in part on the attenuation of the signal strength. In this way, the robot 102 can detect how close to the transmitter it is and thus can detect the position of the robot 102 relative to the transmitter and / or initialization position 104. In some implementations, the robot 102 may determine its position by triangulating the signal strength of a plurality of transmitters. In some implementations, the initialization location 104 may be distinguished by an indication (e.g., notation, symbol, line, etc.) on the floor. When one or more sensors (such as sensor unit 314) of the robot 102 detect an indication on the floor, the robot 102 can detect that the robot 102 itself is located at the initialization position 104.

In some implementations, the camera is placed on the ceiling and the camera can be communicatively coupled to the robot 102 (e.g., via the communication unit 316). In some cases, the camera may be part of the sensor unit 314. The camera can determine the position / posture of the robot 102 through image processing and / or machine learning, and can deliver the position / posture to the robot 102. [ In some cases, the camera recognizes when the robot 102 is in the initialization position 104 through image processing and / or machine learning to determine whether the robot 102 is in the initialization position 104, .

In some cases, the user 102 may position the robot 102 at the starting position 104 by the relationship of the starting position 104 to the surrounding objects, while the user 604 positions the robot 102 by the boundary display position on the floor. Will be detected and registered. As an illustrative example, the robot 102 may detect one or more surrounding objects 512, 546, 548, 550 using one or more sensors 560A-560D, as described with reference to Figures 5B- The initialization position 104 can be detected. In some embodiments, and more specifically, the robot 102 may include one or more peripheral objects 512, 546, 548, and 548 using one or more sensors 560A-560D, as described with reference to Figures 5B- (E.g., point 590, 592, 594, 596, 598, etc.) While in the initialization position 104, the robot 102 may initialize its orientation and position.

In some embodiments, from the initialization position 104, the robot 102 may detect the presence of the robot 102 at the initialization position 104 and / or may detect the relative position of the robot 102 to one or more nearby objects And / or direction. In this embodiment, the robot 102 is configured to detect its own sensor (sensor unit 314, etc.) at least to detect that the robot 102 is in the initialization position 104 and to initialize its orientation and / You can use it partially to sense the surroundings. These sensors can sense the characteristics of the surrounding environment such as objects (goods, walls, etc.), floors, ceilings, people and objects, signs, surfaces, Through the relative position and / or orientation of the sensed object in its vicinity, the robot can obtain the orientation relative to the initialization position.

5B-D illustrate an overhead view of the robot 102 at an initialization position 104, wherein the robot 102 detects its presence at the initialization position 104 and / Direction and / or position of the object. As illustrated in FIG. 5B, the robot 102 may be located in the initialization position 104. The robot 102 may include a body having a plurality of sides such as a front surface 502, a right side surface 508, a left side surface 506 and a rear surface 504. [ In addition, the robot 102 may have a top surface 564 and a bottom surface (not shown). Those skilled in the art will appreciate that the robot 102 may have other sides corresponding to the surface of the robot 102 that may vary in shape (e.g., rectangular, pyramidal, humanoid or other design geometry). For example, the front face 502 may be positioned on a face facing the front of the robot 102, and the front face may be moved in the forward movement direction of the robot 102. The rear surface 504 may be located on a surface facing the rear of the robot 102, and the rear surface is a surface facing substantially the opposite direction to the front surface. The right side 508 may be the right side with respect to the front side 502 and the left side 506 may be the left side with respect to the front side 502. [

The robot 102 may be part of one or more sensors 314 disposed along one or more of the front side 502, the right side 508, the left side 506 and the back side 504 and / (Which may be substantially similar to the sensor of sensor 560A-560D). May be part of sensor unit 314 described in this disclosure and / or may be any sensor). The sensors 560A-560D may include external water soluble sensors. In some cases, each of the sensors 560A-560D may comprise a number of sensors capable of detecting different characteristics of the environment 100. In addition, the robot 102 may have one or more sensors 568A, 568B that may include external water-soluble sensors. If different types of sensors and / or different sensor coverage (e.g., sensor arrangements for detecting a narrower or wider range of ambient environment 100) are desired, then the sensors may be used more than those shown in Figures 5B- It will be appreciated by those skilled in the art that they can be used heavily or at different locations.

The sensors 560A-560D may be mounted on the side surfaces (front surface 502, right side surface 508, left side surface 506, rear surface 504, top surface 564, bottom surface (not shown), and / Or at an angle. The angle may be determined by the object and the range, focal plane, area of interest and / or other characteristics of each of the sensors 560A-560D. As a non-limiting example, a sonar sensor may emit an acoustic signal (e.g., multiple lobe patterns, a fan, or other characteristic form of a sensor) that spreads from a sonar sensor. For example, FIG. 5E illustrates an overhead view of a robot 102 emitting and / or receiving an energy pattern 580B. The energy pattern 580B is exemplary in nature and does not describe the actual waveform or transmission of the signal. However, the energy pattern 580B is an energy that is emitted by the sensor 560B from the front face 104 and / or reflected from the front face 104, and the energy that enables the detection of the object due to the diffusion of the energy pattern 580B . The energy pattern 580B may be a unique energy used depending on the type of sensor 560B. For example, if the sensor 560B is LiDAR, then the energy pattern 580B may be a pattern of a plurality of light waves emitted from the sensor 560B (and, in some cases, later reflected and received) It can be partially shown. If the sensor 560B is a sonar sensor, the energy pattern 580B may be a pattern of sound waves emitted by the sensor 560B (and, in some cases, later reflected and received). When the sensor 560B is a camera, the flash light or ambient light of the sensor 560B can illuminate the object, and the sensor 560B can detect the reflected light. Thus, in some cases, the energy pattern 580B represents the received energy, rather than the emitted energy, when no energy is emitted by the sensor 560B. If the sensor 560B is an infrared sensor or a 3D camera that detects infrared radiation, the energy pattern 580B may be a pattern of infrared light emitted (and subsequently reflected and received) by the sensor 560B. In the case of an infrared sensor, the sensor 560B may use a filter to detect reflected ambient infrared radiation. As another example, the sensor 560B may be a 3D sensor configured to emit and receive energy to sense the environment in three dimensions. Those skilled in the art will appreciate that other sensors may be used and the energy pattern 580B may at least partially represent the characteristic energy emitted, reflected and / or received by the sensor 560B.

Exemplary sonar sensors may be deployed as one or more of the sensors 560A-560D so that the diffusion covers the desired area or range from the robot 102. [ Measurements (such as distance and / or angle measurements) may be performed on sensors 560A-560D or other locations on the robot 102 body, such as the mass center of the robot 102 or other designated locations.

5b, the robot 102 may use one or more sensors 560A-560D to detect an object 512 in the environment and determine the position of the robot 102 relative to the object 512 and / Direction can be approximated. For example, object 512 may be an obstacle (article, wall, etc.). From the object 512, the robot 102 may measure a distance 516 to a point 590 on the object 512, which may be measured in inches, feet, meters or other units of measurement ). ≪ / RTI > In some implementations, the distance 516 can be measured in relative (or non-absolute) units such as ticks, pixels, percentage of sensor coverage, and the like. In some implementations, the distance 516 may be determined for a reference point such as the initialization position 104, the object 512, the sensor 560A-560D, the mass center of the robot 102, or any other determined position x- and y-coordinates. In this case, the x-coordinate may be the distance to the reference point for the first axis, the y-coordinate may be the distance to the reference point for the second axis, and the second axis, which is orthogonal to the first axis, . In some cases, the distance 516 may be measured in three dimensions including the x- and y-coordinates and the z-coordinates described above, where the z-coordinate may be a distance from the third axis to a reference point.

In some implementations, the one or more sensors 560A-560D may measure or approximate a distance 516 to a point 590 of the object 512. [ For example, the sensor 560A may be a sonar sensor capable of measuring the distance by measuring the time difference between the emitted sound wave and the time the sound wave is reflected back to the sensor 560A, The time difference of the sound wave can be converted into the distance using the sound speed.

In some implementations, the one or more sensors 560A-560D may generate a map 700, as will be described later, and the map 700 includes not only the object 512, but, in some implementations, . Distance 516 may be approximated based, for example, at least partially on approximate measurements taken on map 700, by using relative units on map 700 or by converting the relative units of map 700 into absolute distance measurements .

The robot 102 may also approximate its orientation at the initialization position 104. In some implementations, the robot 102 may be positioned relative to a reference point for an initialization position 104, an object 512, sensors 560A-560D, a mass center of the robot 102, or a point at another predetermined location, 514 can be approximated. The angle 514 can be measured in degrees, radians, or other units. In some implementations, the angle 514 may be measured on a 2D plane, such as a horizontal plane (e.g., a rectangular coordinate system or other measurement of the distance 516 described above). In some implementations, additional angles, such as one or more rolls, yaws, and pitch, of the object 512 relative to the robot 102 may be measured.

As an illustrative example, the robot 102 may measure an angle 514 with respect to the object 512. One or more sensors 560A-560D can approximate an angle 514 with respect to object 512, similar to the way in which distance 516 to object 512 can be approximated. For example, the sensor 560A may be a sonar sensor capable of determining the orientation (e.g., angle 514) of the object 512 relative to the sensor 560A based on the angle of the received reflected energy. As described above, in some implementations, one or more sensors 560A-560D may generate a map 700 that may include an object 512. [ The angle 514 may be approximated based at least in part on the approximation taken on the map 700, for example, by using a relative unit on the map 700 or by scaling the relative unit to the measured distance.

In some implementations, the robot 102 may move its position and / or orientation (e.g., distance 516 and / or angle 514) relative to the object 512 and / 302 and associates its position with respect to object 512 and / or point 590 to initialization position 104. [ In this manner, the robot 102 may detect the initialization position 104 and initialize the position for the object 512 and / or the point 590 when returning to the initialization position 104 thereafter. The detection of the initialization position 104 and the initialization of the position may be performed by the mapping and localization unit 312.

FIG. 5C illustrates an overhead view of a robot 102 positioned at a predetermined angle in an initialization position 104. FIG. At this time the sensor 560A of the robot 102 uses the system 512 and method substantially similar to the way in which the sensor 560A described with reference to Figure 5B measured the distance 516 and the angle 514, The distance 524 to the point 511 and the angle 518 of the point 511. [ 5C illustrates that the plurality of sensors 560A-560D independently measure the distance and angle to the object 512. [ For example, the sensor 560B may be positioned at a point of the object 512 using a system and method substantially similar to the way in which the sensor 560A described with reference to Figure 5B measured the distance 516 and the angle 514, The distance 522 and the angle 520 to the first portion 592 can be measured. In this manner, the robot 102 can detect the initialization position 104 and / or initialize the position and / or orientation of the robot 102 with respect to the object 512. [ In some implementations, the robot 102 may determine the position and / or orientation of the robot 102 relative to the object 512 and / or points 591 and 592 (e.g., distances 516 and 522 and angles 514 520) may be written to the memory 302 and the position and / or orientation of the robot 102 relative to the object 512 and / or points 591, 592 may be recorded in the initialization position 104 Associate. Thus, when the robot 102 later returns to the initialization position 104, it may detect the initialization position 104 and / or the position of the robot 102 relative to the object 512 and / or points 591 and 592 and / Direction can be initialized.

5D illustrates an exemplary robot 102 in which a number of exemplary objects 512, 546, 548, 550 are used to detect the initialization position 104 and / or initialize the orientation and / or position of the robot 102. [ Lt; RTI ID = 0.0 > of < / RTI > The robot 102 is positioned at a point 594 of the object 546 using a system and method substantially similar to the way in which the sensor 560A described with reference to Figure 5b measured the distance 516 and the angle 514. [ The distance 554 and angle 542 to the point 596 of the object 550 and the distance 558 and angle 540 to the point 598 of the object 550. [ The angle 544 can be measured. In this manner, the robot 102 may detect the initialization position 104 and may determine the position of the robot 102 relative to one or more objects 512, 546, 548, 550 and / or points 590, 594, 596, Can be initialized. In some implementations, the robot 102 may determine its position and / or orientation (e.g., distance (s)) relative to one or more points 590, 594, 596, 598 of objects 512, 546, 548, 516, 558, 554, 556) and / or angles 514, 540, 542, 544) to the memory 302 and may write one or more objects 512, 546, 548, 550 and / Associates the position and / or orientation of the robot 102 with respect to the points 590, 594, 596, 598 with the initialization position 104. Thus, the robot 102 may then detect the initialization position 104 and initialize the position and / or orientation of the robot 102 when returning to the initialization position 104 thereafter.

Detecting the initialization position 104 using a plurality of objects 512, 546, 548 and 550 may be advantageous in that the robot 102 can position the initialization position 104 more precisely. In addition, using a plurality of objects 512, 546, 548 may provide additional uniqueness to the initialization position 104, which may assist the robot 102 detecting the initialization position 104 and / / RTI > and / or < RTI ID = 0.0 > robot 102 < / RTI >

As the robot 102 measures the distance and angle to the object as described with reference to Figs. 5B-5D, the robot 102 can initialize the external acceptance sensors 568A-568B. The initialization of the sensors 568A-568B may be accomplished by zeroing the sensors 568A-568B, setting the sensors 568A-568B to their initial values, or by setting the current values of the sensors 568A- And storing it. In some implementations, external acceptance sensors 568A-568B may be initialized for a reference point. As an illustrative example, the robot 102 may move the point 590 relative to the point 590 such that the point 590 is at the origin (e.g., (0, 0) in the 2D map or (0, 0, 0) The acceptance sensors 568A-568B can be initialized. Thus, the robot 102 can measure the distance 516 and angle 514 to the point 590 and determine the initial position and / or orientation of the robot 102 relative to the origin. This determination may be performed by the mapping and localization unit 312. [ In some implementations, using distance 516 and / or angle 514, robot 102 may use its trigonometry for a vector (e.g., distance 516 and angle 514) (E.g., (x, y) in a 2D map or (x, y, z) in a 3D map). For example, in some cases, the x coordinate may be the cosine of angle 514 multiplied by distance 516. In some cases, the y-coordinate may be the sine of angle 514 multiplied by distance 516. Other points, such as, but not limited to, one of the points 591, 592, 594, 596, 598 may similarly be used as the origin, and as shown and / or described with respect to Figures 5B- Trigonometry may be used with the corresponding vectors (e.g., distances 516, 518, 522, 558, 554, 556 and / or angles 514, 518, 520, 540, 542, 544). In some cases, multiple origin points may be provided so that a plurality of points (e.g., two or more of points 590, 591, 592, 594, 596, 598) may initialize robot 102. Using multiple origin points would be desirable to generate multiple maps, to provide multiple origin points to select for computational simplicity, to provide a check of the sensor when one or more sensor readings are wrong, and to provide other benefits .

Preferably, the sensors 568A-568B use an odometer to track the movement of the robot 102 (e.g., travel distance and rotation) relative to the initialization of the sensors 568A-568B have. For example, the sensors 568A-568B may include one or more mileage meters (e.g., wheel encoders such as rotary encoders), video odometers, compass, GPS (Global Positioning System) (E.g., a red, green, blue, depth ("RGB-D") camera, etc.) capable of detecting angular rotation of the robot 102. The IMU can include accelerometers, magnetometers, angular velocity sensors, and the like. For example, if the sensor 568A includes ladder, the displacement (and corresponding position) can be determined based on the positional difference of the different images at different times. If an RGB-D camera is used, the location can be determined using scan matching. The sensors 568A-568B may also include one or more mileage meters to measure the distance traveled by the robot 102. [

Returning to method 400 of FIG. 4, at portion 404, the robot 102 creates a path 116 (see FIG. 1B), recording the path 116 and the map 700 of the environment 100 under the control of the user Shown in Fig. 6A illustrates a side view of an example in which an exemplary user 604 controls an exemplary robot 102. As shown in FIG. The user 604 may be a manager, an administrator, or any other person who can use the robot 102. As shown, the robot 102 may be a floor cleaner configured to clean floors of shops, warehouses, office buildings, homes, storage facilities, and the like. Thus, it may have a brush 608 configured to clean the floor below and / or around the robot 102.

The robot 102 may be trained to associate (e.g., perform later) actions and / or actions with position and / or locus on the map 700. For example, brush 608 may be actuated by actuator unit 318, brush 608 may be on / off actuated, and may be raised and lowered by actuator unit 318. The robot 102 may learn the operation of the brush 608 as the user controls the brush 608 while recording the path 716 and the map 700. [ In some implementations, the map 700 may include actuator commands for operation of the brushes 608 in one or more locations and / or trajectories on the map 700 and / or paths 716 therein. In some implementations, the robot 102 may also have one or more squeegees 616 (squeegee). The squeegee 616 may be a piece of rubber, such as a rubber-edge blade, to clean or scratch the floor. Actuator unit 318 may also be used to raise / lower squeegee 616. Thus, the robot 102 can learn the operation of the squeegee 616 by allowing the user to record and control the path 116 and the map 700. In some implementations, the map 700 may include actuator commands for actuation of the squeegee 616 in one or more positions and / or trajectories. Turning the water on and off, splashing water, turning the vacuum on or off, moving the vacuum hose position, moving the arms, lifting and lowering the lift, rotating any sensor of the camera and / or sensor unit 314, And / or any other movement of the scrubber or other robot type, such as any movement required for the robot 102 to perform the action, may similarly be learned.

In some implementations, when an action and / or an actuator command is associated with a position on the map 700 and / or a path 716 therein, during autonomous navigation, the robot 102, These actions and / or actuator commands may be performed. In some implementations, when autonomously navigating, when the behavior and / or actuator commands are associated with a position and orbit and / or an internal path 716 on the map 700, the robot 102 may move in the same direction and / / RTI > and / or < / RTI > actuator commands when passing a position at the same relative time. Thus, in these implementations, the robot 102 does not perform an action and / or actuator command each time it passes a location (e.g., the robot is cycled and passes the same physical location multiple times) (E.g., a location) in a particular direction, or only passes a specific location (s) of the path.

Those skilled in the art will appreciate that even if the robot 102 is a floor cleaner, it can have many different forms. Figure 6b illustrates a side view of an exemplary body shape for a floor cleaner. Although the figures illustrate, in a non-limiting example, various forms of body shapes, but do not limit the robots 102 to any particular body shape or floor cleaner. The exemplary body shape 652 has an upright shape with a small frame that allows the user to push the body shape 652 back and clean the floor. In some cases, the body shape 652 may have power propulsion that can assist the user in cleaning but also allow autonomous movement of the body shape 652. The body shape 654 has a larger structural shape than the body shape 652. The body shape 654 may be powered to move with little or no user force on the body shape 654 except for steering. The user can steer the body shape 654 as it moves. The body shape 656 may include a seat, a pedal, and a steering wheel to allow the user to drive the body shape 656, such as a vehicle, when the body shape 656 is cleaned. The body shape 658 may have a shape larger than the body shape 656 and may have a plurality of brushes. The body shape 660 may have a sitting or fully enclosed area as the user drives the body shape 660. The body shape 662 may have a platform on which the user may stand while driving the body shape 662.

In addition, as described in this disclosure, the robot 102 may not be a floor cleaner. For additional, non-limiting illustration, FIG. 6C illustrates some additional examples of the body shape of the robot 102. FIG. For example, the body shape 664 illustrates an example in which the robot 102 is an upright vacuum cleaner. The body shape 666 illustrates an example in which the robot 102 is a humanoid robot having an appearance substantially similar to a human body. The body shape 668 illustrates an example where the robot 102 is a dron with a propeller. The body shape 670 illustrates an example in which the robot 102 has a vehicle shape having wheels and rooms. The body shape 672 illustrates an example in which the robot 102 is a rover.

Referring to FIG. 6A, the robot 102 may be configured in various ways for control by the user 604. FIG. As illustrated, the user 604 can walk behind the robot while steer- ing the robot 102 using the steering wheel 610. [ In another embodiment, the robot 102 may be a floor-standing floor cleaner (not shown) on which the user 604 may ride on a seat or standing platform to control the robot 102. In some implementations, the user 604 may remotely control the robot 102 with a remote controller, such as a wireless remote device, a mobile device, a joystick, or any other device known in the art. This control may be used to turn left, right, forward (e.g., to use an accelerator pedal or to instruct the robot 102 to advance), backward (e.g., to instruct the back pedal or robot 102 to reverse) , Brush up / down, water on / off, and the like. In some embodiments, the user 604 may control the actuator unit 318 to drive movement of the robot 102, raise or lower the brush, and turn on / off the water. In other implementations, the robot 102 may not be a floor cleaner, but may be any other robot described in this disclosure.

6D shows a top view when the user 604 controls the exemplary robot 102 and the robot 102 senses its surroundings. The robot 102 can detect an object and map the periphery of the robot 102 using one or more sensors 560A-560D and other sensors as the robot travels along the path 116. [ For example, the robot 102 may emit energy waves 580A, 580C. Energy 580B has been described earlier in this specification with reference to Figure 5E and has been disclosed elsewhere throughout the specification. Energy waves 580A and 580C may be substantially similar to energy waves 580B where energy wave 580A corresponds to sensor 560A and energy wave 580C corresponds to sensor 560C.

7A illustrates an exemplary map 700 and path 716 generated by the exemplary robot 102 as the robot 102 travels in environment 100. As shown in FIG. In some implementations, the creation of the map 700 may be performed by the mapping and localization unit 312. The map 700 may include pixels, with each pixel corresponding to a mapped region of the environment 100. The number of pixels in the map 700 may be determined based on the resolution of the map 700. [ For example, the map 700 may be displayed at various display sizes (e.g., 3.5 inches, 10 inches, 20 inches and / or any other diagonal screen dimensions known in the art) and display resolution (e.g., 800 x 600, 1024 x 768, 1360 x 768, 1680 x 1050, 1920 x 1200, 2560 x 1440, 3840 x 2160, or any known display resolution known in the art). The screen displaying the map 700 may be a square, a circle, a triangle, a hexagon, or other shapes and may be non-square. These screens may be part of the user interface unit 322. The map 700 may be substantially similar in layout to the environment 100 and each pixel of the map 700 may approximate the location of the environment 100. [

In some implementations, a pixel of map 700 may have more than one state, and the pixel state represents at least a portion of the characteristics of the location / location in environment 100 represented by the pixel. For example, the pixels of the map 700 may be binary, wherein the first pixel state (e.g., pixel value) represents an at least partially unobstructed (e.g., navigable) The second pixel state at least partially indicates a block (e.g., no navigation) position. By way of example, a pixel value of zero (0) may at least partially indicate a clear position, and a pixel value of one (1) may at least partially indicate a blocking position.

In some implementations, instead of or in addition to the binary states described above, the pixels of the map 700 may have one or more of the following other pixel states: an unknown location (e.g., location / location without information) A pixel state that at least partially displays; A pixel state that at least partially displays a location / location that should not be moved; A pixel state that at least partially represents a portion of a navigation path (e.g., path 716); A pixel state at least partially displaying an area where the robot 102 travels; A pixel state at least partially indicating an area in which the robot 102 has not traveled; A pixel state at least partially displaying an object; A pixel state that at least partially represents a standing water; And / or any other classification of location / location on map 700. [

The pixels of the map 700 may also store one or more values or pixel states. For example, each pixel of the map 700 may store a plurality of values, such as a vector or a value stored in a matrix. These values include values that at least partially represent the position / orientation (e.g., position and / or orientation) of the robot 102 when the position is measured at a point (e.g., pixel) along path 716 . These values may include whether the robot 102 should clean the location / location or other actions taken by the robot 102

The robot 102 may move along a path 116 (shown in Figure 1B) that may be reflected in the map 700 as a path 716. The robot 102 may be displayed on the map 700 by a robot indicator 702 and the position of the robot indicator 702 within the map 700 may be displayed relative to the robot 102 within the environment 100 The location can be at least partially reflected. At each location, the robot 102 travels along a path 116 and the robot 102 moves from an initialization position 104 or another reference point (e.g., objects 512, 546, 548, 550) Or orientation with respect to any other reference point used by the robot 102 during initialization in the initialization positions 104, 105, 591, 592, 594, 596, 598 and / or initialization position 104. This mapping and localization function may be performed by the mapping and localization unit 312. The initialization position 104 may be represented on the map 700 as a mapping position 724. The termination position 114 may be a mapping position For example, wheel encoders (e.g., rotary encoders), video odometers, IMUs (e. G., Wheel encoders) (Including an accelerometer, a magnetometer, an angular velocity sensor, and the like) 568B) can measure or estimate the distance from the robot (e. G., A reference point) to track movement after initialization of the robot in the initialization position 104. As an illustrative example, one or more of the self- 102 may be a wheel encoder that measures or estimates the distance based on the rotation of the wheels of the robot 102. As another illustrative example, The image odometer may be configured to construct an optical flow field (e.g., using the Lucas-Kanade method or other method) using a Kalman filter or projection, The IMU can be used to measure or estimate the position and / or orientation of the robot 102. [0034]

The robot 102 moves the path 716 to the map 700 as the robot indicator 702 proceeds along the map 700 in a manner substantially similar to the robot 102 navigating through the environment 100. [ Can be recorded. Preferably, in some implementations, the map 700 and path 716 are generated together, and the robot 102 maps the environment 100 at substantially similar times and records the path 716. Thus, in some implementations, the map 700 and the path 716 may be formed together in pairs, where each recorded path is stored only with a particular map.

At each location that is part of path 116, robot 102 may change the state of the corresponding pixel on path 716 in map 700 to a pixel state indicating that the pixel is part of the navigation path. Simultaneously, the robot 102 may use one or more sensors 560A-560D, using systems and methods substantially similar to those described with reference to sensors 560A-560D in conjunction with Figures 5A-5E, The position and / or the direction of the robot relative to the robot can be measured. In this manner, the robot 102 may detect and / or measure the position and / or orientation of the robot 102 relative to an object such as a shelf or wall to fill the map 700, The pixel state can be changed based at least in part on detection and measurement.

When the robot 102 detects an object, the robot 102 can use sensors 560A-560D to detect and / or measure the position and / or orientation of the object in a plurality of directions relative to the robot 102 have. At the same time, the robot 102 may estimate the position (e.g., travel distance) and / or orientation of the robot 102 using sensors 568A-568B. As the robot 102 moves within the environment, different objects may fall within the range of the sensor. For example, the sensor 560B, which may be located on the front face 502 of the robot 102, may have a range 704. [ For example, the robot 102 may detect an object from the front surface 502 to the range 704. [ Similarly, sensors 560A, 560C, and 560D can each have a range and can detect objects within these ranges. As the robot 102 detects an object and determines a relative position and / or direction from the robot 102, the robot 102 can display the position of the pixel corresponding to the detected object on the map 700 have. Such a pixel may be changed to a state that at least partially indicates that these pixels correspond to an object (e.g., a blocking position or a pixel state representing an object).

Because the robot 102 fills the map 700 on a pixel by pixel basis, the map 700 may have certain artifacts. For example, a smoothly appearing wall may appear jagged based at least in part on the signal received at the sensor. For example, if the sensor 560A-560D comprises a sonar, lidar or other sensor that depends on the reflectance from the surface of the sound, light, or other elements, there may be variability within the surface . There may also be motion artifacts and other artifacts and / or distortions.

In some cases, sensors 560A-560D may not be able to sense specific areas. For example, the object may interfere with the function of the robot 102 sensing the region, or the region may appear in a blind spot (e.g., a location not covered by the measurement range of the sensor). As another non-limiting example, the box 706 highlights on the map 700 the measurements taken by the robot 102 when the robot 102 turns 708 on the map 700. As the robot 102 rotates, the sensors 560A-560D measure the area indicated by white (e.g., navigation position) by the box 706, but some objects interfere with the sensor's range and enter the box 706 Creating a long, crumbly appearance as shown.

As the robot 102 travels along the path 116 from the initialization position 104 to the end position 114 the robot 102 moves along path 116 and path 116 within the range of sensors of the robot 102 And the surrounding environment 100 of the map 100. FIG. 7B shows an exemplary map 700 once completed. Preferably, the robot 102 may record the mapped path 716 and map the environment of the mapped path 716 of the map 700 to one demo. Thus, the map 700 allows the robot 102 to autonomously navigate the path 116 (or a path substantially similar to the path 116) as a demo.

Other modern systems and methods may require a user to use various demos to upload maps, draw paths to maps, or map the environment. Such systems and methods may be burdensome to the user. For example, such systems and methods may provide a cumbersome and low user experience, although they may perform all steps in a satisfactory manner for those systems and methods that the user will be working with. Recording the mapped path 716 of the robot 102 and mapping it to a demo in the map 700 allows the user to train and / or program the robot 102 with minimal user interaction Can be advantageous. This capability is advantageous in that it is easily adaptable to many environments based on relatively few user demos.

Referring again to FIG. 4, at portion 406, the robot 102 may determine a mapping error in the map 700. This determination can be performed by the map evaluation unit 324. [ Preferably, if the robot 102 autonomously travels along the path 106 after generating the single demo map 700 (e.g., autonomous step 416), then (path 716) Determining whether there is a mapping error in the map 700 (e.g., including a map) may allow the robot 102 to avoid negative consequences of, for example, collisions, errors and / or incorrect or incorrect mappings. If the robot 102 finds that there is a sufficient mapping error in the map 700 and / or that the quality of the map 700 is not good, then the robot 102 may determine that the quality of the map (e.g., via the user interface) (E. G., User 604 or other user) a warning, an alarm, a prompt, and / or other indication that it is not good. In some cases, the robot 102 may send an alert, an alarm, a prompt, or other indication to the user to re-demonstrate the path (e.g., by performing portions 402 and 404 again). Preferably, the robots 102 are prevented from colliding with the obstacles by determining errors prior to autonomous navigation and / or by evaluating the quality of the map 700, and are prevented from stopping due to the mapping of the robots 102 Time can be saved and damage can be prevented.

There are a number of ways in which the robot 102 can detect a mapping error and / or evaluate the quality of the map 700 (including path 716), and each method is implemented singly or in combination. In particular, the presence of all mapping errors or mapping errors does not imply that the map 700 is of poor quality and / or does not mean that it can not be used for autonomous navigation. In fact, the map 700 may have many errors and still be suitable for autonomous navigation. Rather, portion 406 can be used to determine if map 700 is sufficiently vulnerable such that robot 102 can not autonomously navigate or autonomously navigate based at least in part on map 700. The foregoing provides some illustrative examples of how the robot 102 can make such an assessment. In some implementations, when the mapping error is detected and / or the quality of the map 700 is evaluated, the robot 102 may consider at least some characteristics of the error in the map 700. [ Preferably, in some cases, the robot 102 may detect a mapping error and / or assess the quality of the map 700 with little or no input and / or effort by the user 604. [ This gives the user 604 a solid experience of further emphasizing and enhancing the autonomy of the robot 102.

As an illustrative example, in some implementations, the robot 102 may be coupled to any interface for a server, control center, mobile device, and / or user / viewer to verify the map 700 and / The map 700 can be transmitted. The viewer may view the map 700 on a display, such as a screen, a computer monitor, a television, etc., and / or on any display in the user interface unit 322. The viewer may also reply to the robot 102 and this communication may at least partially indicate whether the map 700 and / or path 716 is appropriate for autonomous navigation. In some cases, the robot 102 may be able to receive communications that can transmit the map 700 and / or at least partially indicate whether the path 700 and / or route 716 is suitable for autonomous navigation The communication unit 316 can be used to transmit the map 700. [ In some cases, a user interface (e.g., user interface unit 322) may be on the robot 102, where the user can view the map 700 and / or path 716, And / or path 716 are suitable for autonomous navigation.

As another illustrative example, in some implementations, the robot 102 may retrieve a specially predetermined pattern (e.g., a predetermined error pattern) of the map 700, including path 716, The presence or absence of a predetermined pattern may at least partially indicate a mapping error and / or the quality of the map 700. [ As an illustrative example, when the robot 102 is a floor cleaner that operates in a shop, the robot 102 may include one or more series of approximately parallel objects 108, 110, 112 (which may represent a shelf for displaying merchandise) (As shown in Figures 1A-1C). As shown in map 700, objects 108, 110, and 112 may appear parallel to mapped objects 808, 810, and 812, as shown in FIG. 8A. Therefore, when the robot 102 maps the mapped objects 858, 860, 862 instead, as illustrated in Fig. 8B, the robot 102 can know that the map 700 is defective.

The robot 102 can detect this specific pattern on a pixel-by-pixel or area-by-area basis. In some cases, the robot 102 may use image processing such as segmentation, edge detection, shape recognition, and / or other techniques to identify one or more objects 858, 860, 862 in the map 700. Once objects 858, 860 and 862 are identified, robot 102 may use various methods to determine whether objects 858, 860 and 862 are approximately parallel to other objects of objects 858, 860 and 862 . The robot 102 may then measure the orientation and / or position of objects 858, 860, 862, such as the distance between objects 858, 860, 852 and / or relative angles. Based at least in part on the measured direction and / or position, the robot 102 may determine whether the objects 858, 860, 862 are approximately parallel.

As an illustrative example, robot 102 may use seeding or area growth to define objects 858, 860, 862 (e.g., to find corresponding pixels). Using these pixels, the robot 102 can identify a number of points within the objects 858, 860, 862. As an illustrative example, the robot 102 may identify points 868, 866, 864 in the object 862 and points 890, 892, 894 in the object 862. The robot 102 measures and compares the distances between the points 864, 866 and 868 of the object 862 and the points 890,892 and 894 of the object 860 to determine whether the objects 860 and 862 are approximately parallel Can be determined in part. For example, if the difference in distance between point 866 and point 892 and between point 868 and point 894 is greater than or equal to a predetermined threshold (e.g., a possible deviation from the actual position of the shelf in measurement or approximately parallel , The robot 102 can know that the objects 860 and 862 are not substantially parallel to each other. In some cases, the predetermined threshold value may be stored in the memory 302. If the difference in distance is below a predetermined threshold or the like, then the robot 102 will know that they are almost parallel. Those skilled in the art will appreciate that the robot 102 may similarly calculate differences in distance and distance using points and / or other points other than points 864, 866, 868, 890, 892, 894 in objects 860, 862 . The robot 102 may similarly compare objects 858, 860, 862 and / or may similarly compare any other objects that may be present. If the robot 102 finds one or more substantially non-parallel objects at a location that anticipates parallel objects such as objects 108, 110, 112 and 118, the robot 102 may detect mapping errors in the map 700 and / Or the map 700 is not of good quality. In some cases, the robot 102 may prompt the user 604 to demonstrate the path again (e.g., via the user interface unit 322).

8C illustrates an exemplary mask 870 that may be used to search for a map 700 for parallel objects such as objects 808, 810, and 812. In other exemplary implementations, Mask 870 can be a structural template that can be visualized as a matrix, and each cell in the matrix can represent a pixel group of pixels or maps 700, and their corresponding pixel states. As used in certain applications in the art, mask 870 may also be referred to as a filter. The mask 870 may be stored in memory 302 and / or in a portion of the software configured to process the map 700. In some implementations, the size of the mask 870 may be determined based on the size of the map 700 and objects 808, 810, 812 (e.g., m pixels in the x direction and n pixels in the y direction) As an < RTI ID = 0.0 > mxn < / RTI > For example, the size of the mask 870 may be determined based at least in part on the percentage of the overall pixel dimensions (e.g., 5%, 10%, 15%, 20%, 25% May be predetermined or may be determined based at least in part on known approximate measurements of objects 808, 810, 812. In some cases, the mask 870 may vary in magnitude through repetition of the search method, where the mask 870 begins to search the map 700 as the first size, And then searches the map 700 for a predetermined number of times by searching again for the third size. For example, the mask 870 can be started as a larger mask and become a smaller mask through successive iterations. The size of the mask 870 illustrated in FIG. 8C is for illustrative purposes and may be scaled.

The mask 870 may search the map 700 by sweeping around and around the map 700 and comparing the contents of the mask 870 with the contents of the map 700. [ For example, the mask 870 may be a matrix, and each cell of the matrix may be associated with a pixel state (e.g., an unobstructed (navigable) location, a blocked (e.g., non navigable) A value corresponding at least in part to a location, a location that should not be run, a portion of a navigation path, a ridden location, a non-ridden location, an object, water, and / or any other categories of the map 700 described in this disclosure) Respectively. (E.g., the right upper corner cell, the lower left corner cell, the lower right corner cell, the intermediate cell, or any other cell of the mask 870) of the matrix 700, or any other cell May be aligned with one or more or all pixels. As these cells are aligned with each pixel of the map 700, other cells of the mask 870 may also be aligned with the surrounding pixels of the map 700. [ Each pixel aligned from the map 700 may be compared with a corresponding pixel of the mask 870 to detect similarity between the mask 870 and the area of the map 700 where the pixels are aligned.

As illustrated, mask 870 defines structures 876 and 878 that can at least partially represent parallel objects (e.g., two of objects 808, 810, and 812). A cell (e.g., cell 876) of structures 876 and 878 may have a value that indicates a particular characteristic of the object searched. For example, each of the cells of structures 876 and 878 may have a value that at least partially represents an object of map 700 (e.g., at least partially representing a pixel state for an object in map 700) . Between structures 876 and 878 there may be a structure 880 where the pixels may have values that indicate an obstacle-free location. In this manner, in some implementations, structures 876 and 878 may represent shelves and structure 880 may represent a passage therebetween. Thus, each cell of the mask 870 may have a value that represents the expected pixel of the map 700. The designation of the cells in the mask 870 may reflect the pattern of pixels in which the map 700 is searched. In some implementations, in an iterative search, the mask 870 may rotate and / or change direction. Preferably, this allows the mask 870 to search the map 700 for items that are tilted at an angle and / or allow the map 700 itself to tilt at an angle.

The mask 870 identifies a group of pixels (e.g., having a predetermined matching threshold of 70%, 80%, 90%, or more) that substantially matches the cell values of the mask 870 in the map 700 (E.g., a message, a value or an instruction), and the robot 102 may generate a display (e.g., a message, a value or an instruction) A match can be found and / or the location of such a match can be found. In some cases, if too little matching is found (e.g., based on a predetermined number of expected items to be found), the robot 102 may detect a mapping error in the map 700 and / 700) is not good. In some cases, the robot 102 may detect a mapping error in the map 700 and / or the map 700 is in a good state (e.g., when the mask 870 is configured to identify an undesirable structure) It can be decided not to do so. In either case, the robot 102 may prompt the user 604 to re-demonstrate the path (e.g., via the user interface unit 322).

As another example, in some implementations, the robot 102 may find discontinuities in the map 700 and / or path 716. FIG. 9A illustrates an exemplary path discontinuity 904 between an exemplary path portion 902A and an exemplary path portion 902B of an exemplary mapped portion 900. The mapped portion 900 may be part of the map 700. The mapped portion 900 may include objects 906A-906B and an unobstructed space 908 formed therebetween. Within the obstacle-free space 908, the path is shown as path portion 902A and path portion 902B. There is a path discontinuity 904 between the path portion 902A and the path portion 902B. The path discontinuity 904 may at least partially indicate an error because the robot 102 can move from the path portion 902A to the path portion 902B or vice versa without going to any space therebetween It is because it did not proceed. In some cases, the path discontinuity 904 may cause the robot 102 to move to a position where the robot 102 is mapped, since the robot 102 can pass through the obstacle-free space 908 from the path portion 902A to the path portion 902B without problems It may not be a problem to search for the path 716. [ However, the path discontinuity 904 may at least provide a mapping error and / or a quality of the map 700 (e.g., the map 700 is poor in quality), along with other path discontinuities and / It can be partially shown.

When the robot 102 detects a mapping error and / or evaluates the quality of the map 700 and there are other path discontinuities in other parts of the map 700, the robot 102 may determine the size of the path discontinuity 904 The number of pixels of the path discontinuity 904, the distance, etc.). In some cases, if path discontinuity 904 is greater than a predetermined size threshold (e.g., stored in memory 302), robot 102 may detect a mapping error and / or map 700 It can be determined to be bad. The predetermined size threshold can be measured with an absolute distance measurement using standard units such as inches, feet, meters, or other measurement units (eg, metric, US or other measurement systems), or a percentage of ticks, pixels, (Or non-absolute) measurement unit, such as a range. This predetermined magnitude threshold may be determined based at least in part on at least one of the following factors: signal resolution and / or fidelity (e.g., of sensor unit 314) of the sensor of robot 102; Complexity of the environment 100; An empirical correlation between the path discontinuity for the robot 102 and the mapping error / poor map quality; The ability of the robot 102 to navigate to the path discontinuity 904; And / or other factors. For example, if the signal resolution and / or fidelity of the sensors of the robot 102 is low, then the robot 102 may expect to have a path discontinuity (e. G., Path discontinuity 904) in the mapping , Such path discontinuities can be large in size. The predetermined size threshold may be relatively high since the presence of this path discontinuity may not at least partially indicate a mapping error and / or a poor map quality. In contrast, if the signal resolution and / or fidelity of the sensors of the robot 102 is high, then the path discontinuity 904 may not be predicted, and a small-sized path discontinuity may also cause map errors and / And thus the predetermined magnitude threshold may be relatively low. As another example, a highly complex environment 100 may modify the mapping and localization functions (e.g., mapping and localization unit 312) of the robot 102 and the discontinuity 904 may be expected, The predetermined size threshold may be relatively high. In contrast, the relatively simple environment 100 may not compromise the mapping and localization capabilities of the robot 102, and the path discontinuity 904 may not be expected, and thus the predetermined size threshold may be relatively low have. As another example, if environmental safety is a concern, the predetermined size threshold may be relatively low. As another example, the robot 102 may map a previous map (or a map aggregated on a server) whose map quality (and / or no mapping error) has been independently evaluated (e.g., by a user or another person) Lt; / RTI > The robot 102 may determine a predetermined size threshold to detect a mapping error based at least in part on the discontinuity 904 and / or other path discontinuities and / or to assess the quality of the map 700, The correlation between the sizes of the parts can be considered. As another example, the predetermined magnitude threshold may be based, at least in part, on the ability of the robot 102 to navigate the map 700. When the path discontinuity 904 is greater than a predetermined size threshold, the robot 102 can no longer navigate along the map 700 and thus the robot 102 can detect a mapping error and / 700) may be determined to be poor. In the case of a determination of a detected error and / or poor quality, the robot 102 may prompt the user 604 to re-demote the path (e.g., via the user interface unit 322).

Similarly, the path discontinuity 904 may be one of a plurality of path discontinuities in the map 700. The robot 102 may consider these other path discontinuities. If the number of path discontinuities exceeds a predetermined threshold (e.g., stored in the memory 302), the robot 102 may detect a mapping error and / or determine that the map 700 is bad. For example, the predetermined number of thresholds may be determined based on signal resolution and / or fidelity (e.g., of sensor unit 314) of the sensor of robot 102; Complexity of the environment 100; An empirical correlation between the path discontinuity for the robot 102 and the mapping error / map quality; The ability of the robot 102 to navigate to the path discontinuity 904; And / or other factors. For example, if the signal resolution and / or fidelity of the sensors of the robot 102 is low, the robot 102 may expect to have some path discontinuities (e.g., path discontinuities 904) in the mapping have. The presence of such path discontinuities may not at least partially indicate a mapping error and / or bad map quality, so that the predetermined number of thresholds may be relatively high. In contrast, if the signal resolution and / or fidelity of the sensors of the robot 102 is high, the discontinuity 904 may not be expected and the presence of path discontinuities may at least partially indicate a mapping error and / Therefore, the predetermined number of threshold values may be relatively low. As another example, a highly complex environment 100 may modify the mapping and localization functions (e.g., mapping and localization unit 312) of the robot 102, and the path discontinuity 904 may be expected The predetermined number of thresholds may be relatively high. In contrast, the relatively simple environment 100 may not modify the mapping and localization capabilities of the robot 102, and the path discontinuity 904 may not be predictable, so that a certain number of thresholds may be relatively low have. As another example, if environmental safety is a concern, the predetermined number of thresholds may be relatively low. As another example, the robot 102 may have a prior map (or a map aggregated on the server) in which the map quality (and / or no mapping errors) is independently evaluated (e.g., by a user or another person) have. The robot 102 may determine a predetermined size threshold to detect a mapping error based at least in part on the discontinuity 904 and / or other path discontinuities and / or to assess the quality of the map 700, The correlation between the number of negatives can be considered. As another example, a predetermined number of thresholds may be based, at least in part, on the ability of the robot 102 to navigate the map 700. After determining the predetermined number of thresholds of the path discontinuity substantially equal to the path discontinuity 904, the robot 102 is no longer able to navigate the map 700, so that the robot 102 can detect a mapping error And / or may determine that the map 700 is bad. In the case of a determination of a detected error and / or poor quality, the robot 102 may prompt the user 604 to re-demonstrate the path (e.g., via user interface unit 322).

In some cases, a hybrid threshold value that is used in combination with a predetermined threshold value and a predetermined number of threshold values described above may be used. For example, if the predetermined number of thresholds is exceeded, then the map 700 may determine that it contains a mapping error and / or a predetermined number of thresholds that are determined to be of poor quality may include the number of path discontinuities At least partially. When the mapping error is detected and / or it is determined that the map 700 is of poor quality, the robot 102 may redirect the path to the user 604 (e.g., via the user interface unit 322) You can be prompted to do so.

9B illustrates an exemplary object discontinuity 924 between an exemplary object portion 926A of an exemplary mapped portion 920 and an exemplary object portion 926B. The mapped portion 920 may be part of the map 700. As illustrated, the path portion 922 may not have a path discontinuity. Between the object portion 926A and the object portion 926B, however, there may be an object discontinuity 924 where a portion of the object is not mapped. The object discontinuity 924 may indicate an error because it is likely that it is an unmapped portion of the map portion 924 at the location where it should be mapped. In some cases, since the robot 102 can detect the presence of an object with the sensor when the robot 102 is navigating through the path portion 922, the object discontinuity 924 can be used to navigate the robot 102 It may not be a problem. However, the object discontinuity 924 may itself exhibit a mapping error and / or a bad map with other discontinuities and / or other characteristics of the mapping error.

Similar to evaluating the mapping error and / or evaluation of quality described with reference to FIG. 9A, in evaluating the map 700, the robot 102 may determine the object discontinuity 924 (e.g., number of pixels, The distance of the object discontinuity 924, etc.) and whether there are other object discontinuities elsewhere in the map 700. In some cases, when the object discontinuity 924 is of a size above a predetermined size threshold (e.g., stored in the memory 302), the robot 102 determines whether the map has a mapping error and / Can be determined. For example, the predetermined size threshold may be: signal resolution and / or fidelity of the sensors (e.g., sensor unit 314) of the robot 102; Complexity of the environment 100; Empirical correlation between object discontinuity with the robot 102 and mapping error / map quality; The ability of the robot 102 to navigate to the object discontinuity 924; ≪ / RTI > and / or other factors. For example, if the signal resolution and / or fidelity of the sensors of the robot 102 is low, the robot 102 may expect some object discontinuities in the mapping, and such object discontinuities may be of a larger size. The presence of such object discontinuities may not exhibit, at least in part, mapping errors and / or low map quality, and this predetermined size threshold may be relatively high. In contrast, when the signal resolution and / or fidelity of the sensors of the robot 102 is high, the object discontinuity 924 is unexpected and even small discontinuities can at least partially indicate a mapping error and / Therefore, the predetermined size threshold value may be relatively low. As another example, a highly complex environment 100 may change the mapping and localizing capabilities of the robot 102 (e.g., of the mapping and localization unit 312), and the object discontinuity 924 may be predictable And the predetermined size threshold is relatively high. In contrast, the relatively simple environment 100 may not change the mapping and localizing capabilities of the robot 102, and the object discontinuity 924 may not be expected, and therefore the predetermined size threshold is relatively low. As another example, if environmental safety is of interest, the predetermined size threshold is relatively low. As another example, the robot 102 may have a prior map (or maps gathered on the server) whose map quality (and / or no mapping errors) is independently evaluated (e.g., by the user or others) have. The robot 102 then determines the magnitude of the object discontinuity in determining the predetermined magnitude threshold in detecting the mapping error and evaluating the quality of the map 700 based at least in part on the object discontinuity 924 and other object discontinuities. Can be considered. As another example, the predetermined size threshold may be based, at least in part, on the ability of the robot 102 to navigate the map 700. After the object discontinuity 924 has become larger than a predetermined size, the robot 102 may not be able to navigate the map 700 any further and thus the robot 102 may detect a mapping error and / 700) can be determined to be of low quality. In some instances where an error is detected and / or there is a low quality decision, the robot 102 may then prompt the user to demo the path again (e.g., via the user interface unit 322).

Similarly, the object discontinuity 924 may be one of a plurality of object discontinuities or the like of the map 700. The robot 102 may consider these other object discontinuities. If the number of object discontinuities is above a predetermined number threshold (e.g., stored in the memory 302), the robot 102 detects a mapping error and / or determines that the map 700 is of poor quality. For example, the predetermined number threshold may be a signal resolution and / or fidelity of the sensors (e.g., sensor unit 314) of the robot 102; Complexity of the environment 100; Empirical correlation between object discontinuity with the robot 102 and mapping error / map quality; The ability of the robot 102 to navigate to the object discontinuity 924; And / or other factors. ≪ RTI ID = 0.0 > [0031] < / RTI > For example, if the signal resolution and / or fidelity of the sensors of the robot 102 is low, then the robot 102 may expect some object discontinuity in the mapping (e.g., discontinuity 904). The presence of such object discontinuities may not exhibit mapping errors and / or low map quality, and this predetermined number threshold may be relatively high. In contrast, when the signal resolution and / or fidelity of the sensors of the robot 102 is high, the object discontinuity 924 is not expected and the presence of object discontinuity may at least partially indicate a mapping error and / Therefore, the predetermined number threshold value may be relatively low. As another example, a highly complex environment 100 may change the mapping and localizing capabilities of the robot 102 (e.g., of the mapping and localization unit 312), and the object discontinuity 924 may be predictable And the predetermined number threshold value is relatively high. In contrast, a relatively simple environment 100 may not change the mapping and localizing capabilities of the robot 102, and the object discontinuity 924 may not be expected, and therefore the predetermined number threshold is relatively low. As another example, if environmental safety is of interest, the predetermined size threshold is relatively low. As another example, the robot 102 may have a prior map (or maps gathered on the server) whose map quality (and / or no mapping errors) is independently evaluated (e.g., by the user or others) have. The robot 102 then detects the mapping error and evaluates the quality of the map 700 based, at least in part, on the object discontinuity 924 and other object discontinuities, to determine the magnitude of the object discontinuity Can be considered. As another example, the predetermined size threshold may be based, at least in part, on the ability of the robot 102 to navigate the map 700. After the predetermined number of object discontinuities 924 have become substantially similar to the object discontinuities 924, the robot 102 may not be able to navigate the map 700 any further, And / or determine that the map 700 is of poor quality. In the event of detection of an error and / or low quality of determination, the robot 102 may then prompt the user to re-demote the path (e.g., via the user interface unit 322).

In some cases, hybrid thresholds may be used when used in combination with a predetermined size threshold and a predetermined number threshold. For example, the predetermined number of thresholds determined that map 700 has a mapping error and / or poor quality may be based, at least in part, on the number of object discontinuities exceeding a predetermined size threshold. If the mapping error is detected and / or the map 700 is determined to be of poor quality, the robot 102 may instruct the user 704 to re-demote the path (e.g., via the user interface unit 322) You can prompt.

9C illustrates an exemplary mapped portion 920 having a discontinuity 934 that includes both a path discontinuity and an object discontinuity. The mapped portion 920 may be part of the map 700. The discontinuity 934 may be discontinuous between the path portion 930 and the path portion 932. The discontinuity 934 may also be discontinuous of the object 936. As described with reference to Figs. 9A to 9C, both the path discontinuity and the object discontinuity can at least partially indicate a mapping error and / or a bad map quality. When the robot 102 evaluates the map 700, the robot 102 may consider both the path discontinuities or the object discontinuities, or both, in detecting the mapping errors and / or determining the quality of the map 700.

As another example, in some implementations, the robot 102 may detect items (e.g., paths, obstacles, or other objects) in the map 700 when detecting a mapping error and / Lt; / RTI > can be evaluated. Figure 10 shows an exemplary mapped portion 1000 with overlapping objects 1002, 1004, and 1006. The mapped portion 1000 may be part of the map 700. As shown, objects 1002, 1004, and 1006 may be walls, objects, shelves, etc., detected by robot 102 during creation of map 700. Based at least in part on the measured position and orientation of the robot 102, the robot 102 maps the objects 1002, 1004, and 1006. Since the estimated positions and / or directions of each of the objects 1002, 1004, and 1006 are on top of each other, the robot 102 can determine that there was an error in the mapping. In identifying such overlapping areas, the robot 102 may examine the pixel-by-area or area-by-area map 700. In some cases, the robot 102 may use a mask and / or a filter to find a predetermined shape in the map 700 (e.g., substantially similar to a modified mask 870 to find a given shape) box). The predetermined shape may be based at least in part on the known error of the robot 102 in the mapping, such as previously observed deformations of the object position and / or sensor error.

 The overlap may also be identified, at least in part, by the excessive density of the objects 1002, 1004, and 1006 detected within and / or around the pixel or pixel region. In some cases, the robot 102 can detect the shape of the map 700, that is, the irregularity of the shape. For example, the robot 102 may detect a captured space, such as the space 1008. In some cases, the space 1008 may be a clean, running and / or navigable space. The space 1008 does not generally occur between the objects 1002, 1004, and 1006 because the robot 102 does not approach the space 1008 as it is mapped. Accordingly, the robot 102 can determine that the map 700 has a mapping error and / or is of poor quality when it detects the space 1008. [ As another example, the robot 102 may detect jagged overhangs 1010 and 1012. Shape irregularities allow the robot 102 to determine that there was an error mapping in one or more of the objects 1002, 1004, and 1008 because such overhangs do not normally occur in the environment 100. Thus, based at least in part on the irregularities in the overhangs 1010, 1012, the robot 102 may detect a mapping error and / or may determine that the map 700 is bad.

As another example of an identifiable mapping error by recognizing the overlap, the robot 102 (and / or the path traveled by the robot 102) may be represented in the map as passing through the object. Since the robot 102 will not pass through objects, such an occurrence may at least partially indicate a mapping error.

As another example, the robot 102 may identify the mapping errors and / or the quality of the map 700 by comparing the map 700 with data from at least one of the sensors of the robot 102. For example, in some implementations, the map 700 is generated using, at least in part, one or more sensors 560A-560D and one or more sensors 568A-568B. However, a check on the accuracy of map 700 may compare map 700 with data recorded less than all sensors 560A-560D and sensors 568A-568B. As one illustrative example, one or more sensors 568A-B may determine the odometry of the robot 102. [ The expression of the path of the robot 102 based only on the mileage measurement can be regarded as a map in the running frame. This map in the running frame can be compared to the map 700, for example, using a comparator, subtracting, and / or any other method of comparing the map in this disclosure. The deviation between the running frame and the map in the map 700 may be an arbitrary value determined based at least in part upon a predetermined threshold (e.g., at least partially based on empirical determination of a correlation to 40%, 50%, 60% , The robot 102 may determine that there was a mapping error and / or determine that the map 700 was of poor quality.

As another example, in some implementations, the robot 102 may be configured to run in a closed loop (e.g., the end position is substantially similar to the initial position). It should be noted that the robot 102 is not always able to run in the closed loop. For example, FIG. 1A illustrates a path that does not form a closed loop because initialization position 104 is illustrated as being not substantially the same position as ending position 114. FIG. 11A shows a robot 102 moving in an exemplary closed loop path 1104 where position 1102 is an initial position and an end position. In this case, if the map of path 1104 does not have an initial position and an end position at roughly position 1102, robot 102 may detect a mapping error and / or determine that the quality of the map is not good . In some cases, a predetermined distance threshold may be present (e.g., stored in memory 302). If the mapped initialization and ending positions are not within a predetermined distance threshold (e.g., the distance between the initialization position and the ending position does not exceed a predetermined distance threshold), the robot 102 may detect a mapping error And / or determine that the map is of poor quality. This predetermined distance threshold may be determined based at least in part on the size of the map (e.g., the predetermined distance threshold may be a percentage of the map size), sensor resolution and / or fidelity, and / or other factors.

As another example implementation, the robot 102 may have an uploaded map of the environment stored in the memory 302. The robot 102 can compare the map 700 with the uploaded map. As an example, the robot 102 may use one or more comparators of the map evaluation unit 324 to compare the uploaded map with the map 700 on a pixel by pixel or area basis. In some implementations, the uploaded map and / or map 700 may be resized to facilitate its comparison. If it is found that the map 700 is not similar to a pixel-by-pixel or area-uploaded map, the robot 102 may determine that there is a mapping error and / or that the quality of the map 700 is not good. As a result, the robot 102 may prompt the user 604 to re-demonstrate the path (e.g., the robot 102 may perform the portion 404 again).

In some implementations, a percentage similarity may be calculated between the uploaded map and the map 700, where the percentage similarity at least partially reflects how similar the uploaded map is to the mapping (700). If the percentage similarity falls below a predetermined threshold (e.g., any percentage that at least partially represents a 70%, 80%, 90%, or substantial similarity between the uploaded map and the map 700), the robot 102 May have a mapping error and / or determine that map 700 is of poor quality. As a result, the robot 102 may prompt to re-demonstrate the path through the user 604 (e.g., via the user interface unit 322) (e.g., ) Can be performed again).

In some implementations, the uploaded map may be analyzed for shapes (e.g., shapes of objects or obstacle-free spaces). The map 700 may be analyzed for the same shape to at least partially determine if the same shape is present in the map 700. [ The mask and / or filter may be used for searching in some implementations (e.g., substantially similar to a modified mask 870 to find a shape). If a shape from the uploaded map is not found in the map 700, the robot 102 may have a mapping error and / or determine that the map 700 is of poor quality. As a result, the robot 102 may prompt to re-demonstrate the path through the user 604 (e.g., via the user interface unit 322) (e.g., ) Can be performed again). Similarly, the map 700 may be analyzed for shapes (e.g., shapes of objects or unobstructed space), and the uploaded map may be analyzed to determine if those same shapes exist. In the same manner, if the robot 102 does not find the detected shape from the map 700 in the uploaded map, the robot 102 determines that there is a mapping error and / or that the map 700 is of low quality (E.g., via the user interface unit 322), the user 604 may prompt to again demote the path (e.g., the robot 102 may perform the portion 404 again).

In some implementations, the robot 102 may analyze the map 700 for certain expected characteristics / characteristics of the environment 100. For example, in a grocery store or similar environment, the robot 102 may expect a passage and / or shelf. If the robot 102 does not detect objects that represent passageways and / or shelves, or if it detects too little or too much, the robot 102 may determine that the map 700 may be of poor quality and / Can be included. As another example, there may be some level of expectation for the complexity of the environment. If the map 700 has too much rotation or too little rotation, the robot 102 may determine that the map 700 may be of poor quality and / or may contain mapping errors. As another example, the environment 100 may have an expected size. If the size of the map 700 is too large or small, the robot 102 may determine that the map 700 may be of poor quality and / or may contain mapping errors. In any of the foregoing cases where the map 700 does not have certain expected characteristics / characteristics of the environment 100, the robot 102 may be configured to map the user (e.g., user 604) May prompt the user to verify map 700. < RTI ID = 0.0 > Therefore, the robot can transmit the map to the server and receive the verification of the quality of the map.

In some implementations, a machine learning algorithm may be used that learns that robot 102 (e.g., controller 304 of robot 102) learns to identify good maps and bad maps. For example, the robot 102 may have a library of maps identified (e.g., manually labeled or mechanically labeled) as good maps and bad maps. The robot 102 may learn to associate characteristics determined through its library with robotic 102 representing a good map or bad map, using supervised or non-supervised algorithms known in the art. Thus, if the robot 102 identifies the map as a bad map, the robot 102 may determine that there is a mapping error and / or that the map 700 is of poor quality (e.g., the user interface unit 322) (E.g., robot 102 can perform portion 404 again). ≪ RTI ID = 0.0 >

In some situations, the robot 102 may also correct errors in the low quality map 700. For example, in some cases, the difference between the initialization position and the ending position may be determined in some cases where the robot 102 does not travel precisely in a closed loop (e.g., the closed loop path 1104) Can be used to correct mileage measurements. For example, the robot 102 may determine the difference between the initial position and the end position, and determine how the difference represents a degree of deviation from the actual distance. Therefore, the robot 102 can adjust the recorded path in consideration of the determined deviation.

As another example, some mapping errors may result in a pattern that can be associated with at least a portion of the corrected map, which may be a version of the map 700 in which the robot 102 corrects one or more errors. FIG. 11B illustrates an example in which the exemplary robot 102 associates an exemplary mapping error with an exemplary modified path 1108. FIG. For example, the map 700 may comprise a series of substantially similar shaped drift paths, such as mapped paths 1106A-1106N, where N is an arbitrary number of mapped paths 1106A-1106N, . The robot 102 may determine that such drifted mapped paths 1106A-1106N may at least partially indicate that the user 604 is continually and repeatedly navigating the same path. As a result, the robot 102 can correct the mapped paths 1106A-1106N to the mapped path 1108, which indicates that the user 604 is navigating the same path repeatedly. If the map 700 includes mapped paths 1106A and 1106N, the robot 102 may correct the mapped paths 1106A-1106N to the mapped path 1108 of the map 700. [ Similarly, there may be other error patterns (e.g., drift and / or other errors) whose identities may be programmed into the robot 102 to enable the robot 102 to automatically correct them. Therefore, the robot 102 can correct the error of the map 700. [

The robot 102 may also learn to use machine learning to correlate errors with corrections of these errors. For example, the robot 102 may store the faulty map in the memory 302 and / or on the server. By way of illustration, in some cases, the user 604 may first demonstrate the path. There may be a mapping error in the generated map of the path and surrounding environment. If a mapping error is encountered, the user 604 may remap the environment and / or path. Thus, the robot 102 may have a low-quality version (e.g., with a mapping error that prevents successful navigation) and a non-low-quality version (e.g., without mapping errors that prevent successful navigation). The robot 102 may then associate at least some of the low quality maps with the corresponding portions of the remapped versions that are not low quality. Based on one or more substantially similar associations, the robot 102 can learn to identify the mapping errors that have occurred and learn to generate at least a portion of the corrected map once it recognizes the mapping errors.

4, after the teaching step 414, the robot 102 may enter the autonomous step 416. [ At portion 408, the robot 102 may detect the initialization position 104 and initialize the position and / or orientation of the robot 102. [ In some implementations, the user may operate the robot (not shown) by driving the robot 102, the remote control robot 102, the steering robot 102, the pushing robot 102 and / or any other control that drives the actuator unit 318 102 to the initialization location 104. [ In some implementations, the robot 102 may autonomously return to the initialization position 104. For example, the robot 102 may store the location of the initialization location 104 (e.g., as described above with reference to Figures 5B-5E) in memory 302 and return to that location .

In some implementations, the robot 102 may be configured to detect the initial position 104 of the system and method used to detect the initialization position 104, as well as in the portion 402 described with reference to Figures 5B-5E, The initialization position 104 can be detected in a similar manner. In some cases, when the robot 102 returns to the initialization position 104 in the portion 408, the position 102 of the robot for, for example, one or more of the objects 512, 546, 548, 550 is stored in the memory 302 (E.g., from portion 402). The robot 102 may determine that the robot 102 is in the initialization position 104 if it detects that the robot 102 is in the same relative position to one or more of the objects 512, 546, 548, In some implementations, the robot 102 may detect that the user is in the initialization position 104 based at least in part on where the robot 102 is stopped. As such, it can be assumed that the user has stopped and subsequently selected the path, as described with reference to portion 410, as the initialization location 104. In some implementations, a transmitter (e.g., a transmitter that transmits communications using RFID, NFC, BLUETOOTH TM, wireless transmission, radio frequency fields, and / or any other communication protocol described in this disclosure) May be at or substantially in proximity to the location 104. The robot 102 may detect that the robot 102 is in the initialization position 104 when it detects that the robot 102 is on top of the transmitter or is substantially in proximity to the transmitter. In some cases, the transmitter may have an operable range that can detect communication from the transmitter only when the robot 102 is in the starting position. As an illustrative example, the transmission range of the NFC may be 10 centimeters or less. Thus, when the robot 102 receives the transmission over the NFC, the robot 102 can detect that it is located at the initialization position 104. [ In some implementations, the robot 102 can receive transmissions from the transmitter and calculate the distance to the transmitter based at least in part on the attenuation of the signal strength. In this manner, the robot 102 can detect how close to the transmitter it is, and thus the position of the robot 102 relative to the transmitter and / or initialization position 104. In some implementations, the robot 102 may determine its position by triangulating the signal strength of a plurality of transmitters. In some implementations, the initialization location 104 may be separated by an indication (e.g., an indication, a symbol, a line, etc.) on the floor. When one or more sensors (e.g., sensor unit 314) of the robot 102 detect a sign on the floor, the robot 102 is positioned such that the robot 102 is positioned at the initialization position 104 Can be detected.

At portion 410, the robot 102 may select a recorded path to autonomously navigate. In some implementations, the selection of a recorded path (e.g., path 116) by the robot 102 may be based, at least in part, on user input. For example, a user may select an input 572 on the user interface 500 (shown in FIG. 5A) on the display 576, where the input 572 allows the user to select the recorded path of the robot 102 Can be selected. After selecting the input 572, the interface 1200 shown in FIG. 8 may appear. 12 illustrates an exemplary interface 1200 that may be used for path selection. Interface 1200 may present a plurality of paths for selection that are displayed as selectable inputs 1202A-1202F. The user may select one of the selectable inputs 1202A through 1202F via a touch (e.g., if the display 576 includes a touch screen) and / or via any other input mechanism of the user interface unit 322 . For example, in some implementations, the input 1202F may correspond to a mapped path 716 learned by the robot 102. When the user selects input 1202F, the robot 102 selects map 700 and mapped path 716 (which is based on a user demo of path 116) based at least in part on the user's selection .

In some implementations, the robot 102 may automatically select the recorded path based on the initialization position detected in the portion 408. For example, the initialization location 104 may be associated with only the demoted path 116 (or may be mapped to the mapped path 716). Similarly, the robot 102 may have other initialization locations associated with other demoted paths. Advantageously, having a plurality of initialization positions allows the user to demonstrate and enable the robot 102 to autonomously move through various paths. In addition, by allowing the robot 102 to automatically select the recorded path based on the initialization position, the robot 102 can start autonomous travel more quickly with a minimum of additional user input.

Referring again to FIG. 4, at portion 412, the robot 102 may autonomously travel along the recorded path selected in portion 410. For example, the robot 102 may autonomously travel using the map 700 and the mapped path 716.

In a subsequent path 716, the robot 102 is at least capable of processing at least one of the data from the map 700, path 716, and sensors 560A-560D and sensors 568A- Unit 326. < / RTI > The sensors 560A-560D may enable the robot 102 to sense an object therearound, as illustrated and described herein with reference to Figure 6D and elsewhere herein. In this manner, the robot 102 can navigate based at least in part on the detection of the map 700 and nearby objects, and the robot 102 can avoid the object being detected. For example, such an object may be a temporary and / or dynamic change to a temporary item or a temporary item and / or environment. In addition, the detection of a nearby object can enable the robot 102 to localize itself on the map 700 based, at least in part, on the determination of the position of the object robot 102 to detect on the map 700. [

In addition, the robot 102 may be configured to detect at least a portion of a distance-measuring sensor (not shown) to at least partially determine its position / posture (e.g., distance and / or direction) relative to the origin as described at least with reference to Figures 5B- (568A-568B). By using one or more of at least one map 700, a path 716, sensors 560A-560D and sensors 568A-568B, the robot 102 may be moved along path 106, May autonomously travel through path 126 as illustrated in FIG. 1C, or at least via autonomous path through environment 100 or any other environment, using method 400.

In addition, while autonomously traveling along path 106, the robot 102 may be trained during the portion 404 and / or may be trained, such as the brush 908 and / or squeegee 616, Various mechanisms on the robot 102 can be operated. Moving the vacuum hose position, moving the arms, lifting / lowering the lift, rotating any of the sensors in the camera and / or sensor unit 314, and / or moving the vacuum hose, The actuation of the learned behavior of the scrubber or other robotic device, such as any movement the robot 102 wishes to perform, may similarly be performed.

FIG. 13 illustrates an exemplary method 1300 of operating the exemplary robot 102. Portion 1302 includes detecting a first position of the robot at an initialization position. Portion 1304 includes generating a map of the navigable path and environment during the demonstration of the navigable path to the robot starting from the initialization position. Portion 1306 includes detecting a second position of the robot at an initialization position. Portion 1308 includes the robot autonomously navigating at least a portion of the navigable path from the initialization position.

As used herein, a computer and / or computing device may be a personal computer ("PC") and a microcomputer, a desktop, a laptop or a mainframe computer, a workstation, a server, a personal digital assistant A computer, an embedded computer, a programmable logic device, a personal digital assistant, a tablet computer, a mobile device, a portable navigation device, a J2ME mounting device, a mobile phone, a smart phone, a personal unified communications device or entertainment device, But is not limited to, any other device that processes an incoming data signal.

As used herein, a computer program and / or software may include any sequence that performs a function, or a human or machine recognizable step. Such computer programs and / or software may be implemented in a variety of ways, including, for example, C / C ++, C #, Fortran, COBOL, MATLABTM, PASCAL, Python, assembly language, markup languages (e.g. HTML, SGML, XML, VoXML) Including an object oriented environment such as Common Object Request Broker Architecture (JRE), JAVATM (J2ME, Java Beans, etc.), BREW (Binary Runtime Environment), and the like.

As used herein, a connection, a link, a transmission channel, a delay line, and / or a radio may include a causal link between any two or more entities (physical or logical / virtual) that enable the exchange of information between entities.

While particular embodiments of the present disclosure are illustrated by the steps of a particular step of the method, it will be appreciated that these descriptions merely illustrate the broad methodology herein and may be modified as desired by the particular use. At certain stages, it may be unnecessary or selectively displayed under certain circumstances. In addition, certain steps or functionality may be added to the disclosed embodiments, or the order of execution of two or more steps may be substituted. All such modifications are considered to be included within the disclosure disclosed herein and claimed.

While the above description has shown, described, and pointed out novel features of the present disclosure as applied to various implementations, it will be understood that various omissions, substitutions and changes in the form and details of the devices or processes shown may be made. The foregoing description is directed to the best mode currently contemplated for carrying out this disclosure. These descriptions are not to be construed as limiting in any way, but rather as illustrative of the general principles of the disclosure. The scope of the present disclosure should be determined with reference to the claims.

While the present disclosure has been shown and described in detail in the drawings and foregoing description, such illustration and description are to be regarded as illustrative or exemplary and not restrictive. The present disclosure is not limited to the disclosed embodiments. Modifications to the disclosed embodiments can be understood and effected by those skilled in the art upon studying the drawings, the disclosure, and the appended claims when performing the claimed disclosure.

The use of a particular term when describing certain features or aspects of the present disclosure means that the term is redefined to be limited to include any particular characteristic of features or aspects of the present disclosure to which the term relates And should not be understood as doing. As used herein, terms and phrases, and variations thereof, particularly in the appended claims, should be interpreted as open, as opposed to limiting, unless explicitly stated otherwise. As used in the preceding examples, the term "including" should be interpreted to mean "including without limitation", "including but not limited to", and the like; As used herein, the term " comprising "is synonymous with " comprises," " comprising "or" characterized ", and does not exclude additional elements or method steps that are inclusive, The word "having" should be interpreted as having at least, and the term " such as "should be interpreted as" Quot; is to be interpreted as "including but not limited to" The word " exemplary "is used to provide an exemplary illustration of the item being described, rather than a comprehensive or restrictive list of the item, and should be interpreted as" as an example, but without limitation, " The terms " known ", " normal ", "standard ", and similar terms are to be construed as limiting the described item to items available in a given time period or at a given time, Shall be understood to include known, normal or standard techniques, which may be known or known in the future or at any time in the future; The use of terms such as "preferably," "preferred," "desired," or "desirable," and the like, means that certain features are inherently important or essential, Or functionality of the present invention, rather than as being implied to be critical to the functionality or functionality of the device. Likewise, a group of items linked to a conjunction "and" shall not be understood to mean that each and every one of those items must be in the group, and unless otherwise stated, "and / or" I must understand. Similarly, an item group associated with a conjunction "or" should not be understood as requiring mutual exclusivity between the groups, and should be understood as "and / or" unless otherwise specified. The terms "about" or "roughly" and the like are synonyms and values modified by the term are used to denote a range of interest associated therewith, wherein the range is ± 20%, ± 15%, ± 10% ± 1%. The term "substantially" is used to indicate that the result (e.g., measured value) approaches the target value, where close is within 80% of the result, within 90% of the value, Or within 99% of the value. Also, as used herein, "defined" or "determined" may include values, conditions, thresholds, measurements, etc., determined in a "predefined" or "predetermined" and / or other manner.

Claims (26)

17. A non-transitory computer readable storage medium having stored thereon a plurality of instructions,
Wherein the instructions are executable by the processing device to operate the robot,
Wherein the instructions, when executed by the processing device, cause the processing device to:
Detect a first arrangement of the robots in an initialization position;
Generate a map of the navigable path and the environment during the demonstration of the navigable path from the initialization location to the robot;
Detect a second configuration of the robot at the initialization position; And
Wherein the robot is configured to autonomously navigate at least a portion of the navigable path from the initialization position. The method according to claim 1,
Wherein the instructions further cause the processing device to: when executed by the processing device, cause the processing device to evaluate the map generated for errors, and to cause the robot to navigate the at least partially- ≪ / RTI > requesting the user to re-demote the path. The method of claim 2,
Wherein the errors comprise at least one of a discontinuity of the navigable path in the map and a discontinuity in the environment in the map. The method according to claim 1,
Wherein the instructions further cause, when executed by the processing device, to cause the processing device to provide instructions to avoid obstacles that are temporarily placed on the robot while autonomously navigating the navigable path. Possible storage medium. The method according to claim 1,
Further comprising instructions configured to cause the processing device to receive a selection of a navigable path from a user interface. The method according to claim 1,
Wherein the robot is a floor cleaner. The method according to claim 1,
Wherein the generated map includes an indication at least partially indicative of an action performed by the robot on the navigable path. The method of claim 7,
The action is a floor clean, non-transitory computer readable storage medium. A method of operating a robot,
Detecting an initial placement of the robot at an initialization position;
Generating a map of the navigable path and the environment during the demonstration of the navigable path from the initialization position to the robot;
Detecting the next placement of the robot at the initialization position; And
And causing the robot to autonomously navigate at least a portion of the navigable path from the initialization position. The method of claim 9,
Evaluating the map generated for errors; And asking the user to demote the navigable path back to the robot based at least in part on the errors. The method of claim 9,
Wherein the demo comprises receiving control signals from a user. The method of claim 9,
Wherein generating the map of the navigable path and the ambient environment further comprises sensing the ambient environment with a three dimensional sensor. The method of claim 9,
Wherein the step of autonomously navigating the robot further comprises receiving a selection of the navigable path from a user interface. The method of claim 9,
Wherein the step of allowing the robot to autonomously navigate comprises navigating using the map of the navigable path and the environment. The method of claim 9,
Further comprising mapping the behavior performed by the robot on the navigable path onto the generated map. 16. The method of claim 15,
Wherein the action includes cleaning the floor. A non-transitory computer readable storage medium having stored thereon a plurality of instructions,
The instructions may be executed by a processing device to cause the robot to operate,
Wherein the instructions are configured to cause the processing device to generate a map of a navigable path and environment when starting from an initialization location and demodulating the navigable path to the robot when executed by the processing device, Computer readable storage medium. 18. The method of claim 17,
Wherein the generated map further comprises an indication that at least partially represents an action performed by the robot on the navigable path. 19. The method of claim 18,
The behavior cleaning the floor. ≪ RTI ID = 0.0 >< / RTI > 18. The method of claim 17,
Wherein the demo of the navigation path is a computer simulation. In the robot,
A mapping and localization unit configured to generate a map of the navigable path and the environment while demodulating the navigable path from the initialization position to the robot; And
And a navigation unit configured to autonomously navigate the robot using the map. 23. The method of claim 21,
Wherein the navigation unit is further configured to determine not to autonomously navigate at least a portion of the navigable path. 23. The method of claim 21,
Further comprising a sensor unit configured to generate sensor data that at least partially displays objects within a sensor range of the robot, wherein the navigation unit is further adapted to autonomously navigate based at least in part on the sensor data generated Configured, robot. 23. The method of claim 21,
Further comprising a first actuator unit configured to actuate a brush. 27. The method of claim 24,
Further comprising a processor configured to associate a position of the map with an actuation of the first actuator unit. 23. The method of claim 21,
Further comprising a communication unit configured to communicate with a server, wherein the robot transmits the map to the server and receives a verification of the displayed quality of the map.

Images